Ice cream for breakfast? Well, sure, why not, especially when it's a wholesome and nutritious high-protein ice cream you make in your Ninja Creami. It might even be healthier than the breakfast you had this morning.

There are several different models of the Ninja Creami, including the Swirl, which I used. You can read my Ninja Creami Swirl review here. Any model will do for making delicious and healthy breakfast 'ice cream', the Swirl can just make it into soft serve if you like. You must freeze your ingredients in the provided container for 24 hours, so be sure to plan a day ahead.

Protein shakes

I've made a number of different protein-forward Creami recipes, but it might take some experimentation to find the ones you like the best. The simplest option is just to freeze a protein shake (or your milk of choice plus protein powder.) If those taste too plain or icy, you can doctor them with extras like vanilla extract, cacao powder, xanthan gum, pudding mix, collagen powder, cottage cheese, or fruit. You can process them on the CreamiFit setting if your Creami has that; if not, the Light Ice Cream setting will work. Personally, I’m not a big fan of protein powders and shakes. But if you do like them, you’ll like them even better as ice cream.

Just yogurt

Another very simple high protein option is to simply freeze a tub or two of your favorite flavored yogurt. That won't need any doctoring to produce a perfectly textured frozen yogurt treat. I used the Frozen Yogurt setting and sometimes also put it through the Swirl mechanism for a true frozen yogurt shop experience. Of course you can skip that step if you have a regular or Deluxe Ninja Creami and not the Swirl model. Either way, you'll probably want to experiment with mix-ins and toppings. Granola, fruit, and/or a drizzle of nut butter make nice breakfast frozen yogurt toppings. I’ve frozen a number of yogurts; my favorites have been apple pie, tiramisu, and creme brulee.

High protein base

Honestly, I have not yet found the perfect 'healthy vanilla Creami base', but I tried this concoction of vanilla greek yogurt, skim milk, and cottage cheese. Sure, it was healthy, and the texture was great, but it didn’t taste like much. The mangoes on top helped! There are so many recipes online, I’ll have to keep searching.

Simple frozen fruit sorbets

I often eat chia pudding for breakfast, a simple concoction of chia seeds, almond milk, and Greek yogurt. I recently made my usual chia pudding, but instead of topping it with fruit, I topped it with seeds, granola, and a few scoops of simple Ninja Creami sorbet. The one-ingredient sorbet was simply a can of peaches in juice, dumped into a Ninja Creami container. I processed it on the Sorbet cycle. This is delicious and works with any kind of fruit. Pears, mangoes, and pineapple are particular favorites. Of course, you can just make simple fruit sorbet for a light, refreshing breakfast.

Açai bowls

I made my own açai bowl with store-bought frozen açai cubes. I blended them with some fruit, milk, and sweetener and used the Lite Ice Cream setting. I topped mine with fruit, granola, hemp seeds, and almond butter, but you could also add coconut flakes, honey, yogurt, other seeds, nuts, or anything you like. This was tasty.

Yogurt, milk, and fruit

A simple cherry vanilla Creami, basically just equal parts milk of choice, vanilla yogurt, and cherries came out quite nicely. You could do this with any kind of fruit you like.

Just the tip of the icecreamberg

I keep thinking of more recipe ideas I want to try, there are so many breakfast-appropriate Ninja Creami recipes out there. And if you want to make a real full-on whole-milk-cream-egg-sugar ice cream and eat that for breakfast? No judgement here. Life is short — eat dessert first!

Peter Thiel is among the most prominent figures in the technology industry, having co-founded PayPal in the 90s and Palantir a few years later. Although he's since stepped back from running technology companies, he plays an active role in business, finance – and indeed in politics.

Minimum government, maximum freedom

Writing an essay for the Cato Unbound journal nearly 20 years ago, Thiel outlined his political thinking as one that prioritizes the individual – and individual freedoms – above the needs or demands of the majority.

In this essay, he explained how he was disillusioned with democratic politics and that it had become an antagonist to "freedom" in the form of free markets and, specifically, the new generation of tech moguls who were in the process of expansion.

The piece, which ran in excess of 2,000 words, courted plenty of attention, as well as controversy, especially given its timing – with society still reeling from the 2008 banking crisis, and concerns over the widening wealth divide heightening.

Democracy dies in darkness

Many incidents and actions in the technology world can be traced back to the ideas at the heart of this essay, for example, the breakdown of press freedoms, specifically in the big tech takeover of the creative industries and journalism (see the Washington Post example).

Modern tech companies are also increasingly involved in politics, with Elon Musk, for example, increasingly (whether deliberately or inadvertently) using Starlink to wield influence during geopolitical conflicts.

Despite Thiel's proclamation to transcend democratic politics via media such as cyberspace or cryptocurrencies, many tech moguls – including the Palantir co-founder – have also made concerted efforts to involve themselves in this process. Musk's direct involvement in the second Trump administration is the most prominent example, but Thiel himself has handpicked candidates to stand in elections, including JD Vance.

Full spoilers follow for The Bear season 5, including its finale.

I love The Bear. It's one of my favorite TV shows of the past decade. And, while I understand the criticism leveled at its third and, to a greater extent, fourth season, I don't necessarily agree with all of it.

Nevertheless, when it was announced that The Bear season 5 would be the FX TV Original's last hurrah, I'll readily admit that I had a feeling in the pit of my stomach. The reason? Other shows I've enjoyed that ended recently, including Stranger Things and The Boys — coincidentally, they also stopped after five seasons — failed to stick the landing with their finales.

I feared that The Bear would go the same way. You can probably sense my relief through the screen, then, that its series finale not only delivered a fittingly bittersweet conclusion to its overarching narrative, but avoided the pitfalls that its TV peers blindly stumbled into.

One of the biggest issues — among many other problems that I don't have the time nor space to cover here — that befell the hit Netflix and Prime Video shows' finales was that they didn't utilize their time well.

Where The Boys season 5 was concerned, the Amazon TV series' rushed ending gave the impression that it simply couldn't wait for its end credits to roll. In sharp contrast, Stranger Things season 5's finale was dragged out to the point where I was begging for its final credits sequence to begin.

Admittedly, The Bear season 5's pacing isn't perfect, but it uses its time more efficiently and wisely. Indeed, part of its success comes from how this season's plot is structured.

Unlike past seasons, The Bear's fifth installment is 24-esque in its storytelling approach. Vis-à-vis its eighth and final chapter notwithstanding, it takes place over the course of a single day.

In episodes 1 to 7, we follow the titular restaurant's ragtag but lovable crew as they navigate what could be their final-ever service, due to the eatery's lack of funds.

That possible scenario ends up being the least of their problems. A massive, unrelenting thunderstorm not only floods the restaurant but also results in the next food delivery being canceled, meaning the staff can only serve what little stock they have left. Throw an overbooked guest list and storm-related traffic issues in for good measure, meaning their table turnover has to be on point, and it's no surprise that tensions and tempers are soon running at an all-time high.

However, rather than do what past entries have done by slowly but purposefully building the pressure until things boil over, The Bear 5 deliberately shows how its eccentric — and, depending on who we're talking about, emotionally damaged — cast of characters have evolved over the five seasons.

A great example of this happens midway through the penultimate episode, titled 'Caramel'. Here, the storm's impact on the restaurant's power supply makes the lights flicker and leads to Carmen 'Carmy' Berzatto dropping a lamb rib-based dish for one of the evening's most important diners.

Previous seasons would've seen Carmy completely spiral and take his anger out on his colleagues, which would've likely resulted in a verbal and/or physical altercation with occasionally confrontational associate — oh, and long-time family friend — Richard 'Richie' Jerimovich.

This time, though, Richie and Carmy's former protégé, Sydney 'Syd' Adamu, calmly prevents him from overreacting, keeping what's already been a nightmarish day on track.

That's just one of myriad instances this season I could point to that show how far these individuals, and the collective as a whole, have grown since the series first premiered on Hulu and Disney+. Compare that to Stranger Things and The Boys, whose approach to individual character development wasn't always on point, and, from a character study perspective, The Bear is — if you pardon the pun — a different beast altogether.

Speaking of character progression, each member of The Bear's primary ensemble gets their moment in the spotlight, as well as a fitting conclusion to their personal arcs.

And when I say everyone, I mean everyone. With Carmy's blessing and Uncle Jimmy's backing, Ebraheim finally starts to turn The Original Beef of Chicagoland into a franchise. Meanwhile, Sydney reaffirms her trust in Tina by appointing her as the restaurant's new chef de cuisine.

Furthermore, Marcus starts to make peace with his estranged father and proves that his recent award win wasn't a flash in the pan. Natalie, aka 'Sugar', and the Berzatto family matriarch Donna's previously strained relationship continues on an upward trajectory. Heck, even Gary, aka 'Sweeps', gets a genuinely crowd-pleasing moment; after the storm ruins the 1997 and 1999 year labels on two bottles of red wine, he correctly guesses the right one – using his wine-based powers of deduction – to serve to a diner who happens to be a wine connoisseur.

The biggest wins are reserved for Carmy, Syd, and Richie, though.

Richie is offered the chance to attend an international hospitality summit in Japan. He's not only accompanied by Jessica on the trip, but showrunner Christopher Storer also gives fans what they've been wanting for this pair by ending the 'will they, won't they' subplot and establishing their romantic feelings for each other.

As for Syd and Carmy, Syd fully establishes herself as The Bear's new head chef, while Carmy quits the food business entirely and, putting his love and skill for drawing to good use, becomes an architectural intern.

However, the biggest cause for celebration is that, almost four seasons after they first set their sights on earning a Michelin star, they don't get one, but two.

Like so many other moments in season 5 – and after watching this duo endure hardship after hardship throughout the series – the emotional release I felt once they processed their achievements and shared a joyful hug was euphoric. It's the sign of great writing when you genuinely care about a bunch of fictional characters, and I certainly lost count of how many times I fought back tears, clapped, and fist-pumped the air in The Bear's final two episodes as its crew individually and collectively triumphed against the odds.

Stranger Things 5 and The Boys 5 occasionally struck an emotional chord, and there's no denying that they have their fair share of impactful moments. Held up to the light against another tale centered on a group of underdogs fighting the impossible, though, and it's clear The Bear just does everything better in a concluding chapter whose every bite is delightfully delicious.

Though it ends in magical albeit bittersweet fashion, The Bear doesn't wrap up every storyline in a neat little bow, nor give every character the perfect send-off, but I suppose that's the point.

As a fixture of the Chicago food scene, the eponymous eatery doesn't end simply because Carmy is no longer a part of it — its story, and the lives of those past and present who were involved in it, continue whether we have a seat at the table to see what happens next or not. As Lucca reminds Marcus towards the end of its finale: "The Bear has something that no other place has... family" — and he's right. If I had to choose a fictional found family to be a part of out of the previously mentioned shows, I'd choose The Bear every time.

All five seasons of The Bear are out now on Hulu (US) and Disney+ (everywhere else), aka two of the world's best streaming services.

Whether AI will replace existing jobs by automating the functions of human employees, or somehow lead to new jobs in the future, is debated to death. Regardless, the AI buildout and rollout is now underway. And in an ocean of AI pessimism, Nvidia's CEO Jensen Huang is relatively bullish on the future of human labour.

Software development in the age of AI

Speaking at GTC 2026, the Nvidia chief attempted to bulldoze over pessimistic views that AI will lead to job losses by arguing that productive companies keen to grow will wield AI as a productivity tool.

By doing so, he suggested, companies can become more productive rather than scale back on staff like engineers altogether. Tech like agentic AI, similarly, may need massive teams of people to manage, configure and operate.

Huang isn't alone in this view, with Gartner even suggesting that after the temporary disruption caused by the AI rollout, the technology will be responsible for the creation of more jobs in the future.

The white-collar cull

This thinking defies much of the received wisdom that has arisen about AI's impact on the labor market – and really stems from parallels drawn with the industrial revolution in the 1800s, when critics – known as the Luddites – levied the same charges against the emerging technologies of the time.

Despite Huang's optimism, the cold, hard reality in the labor market of today – rather than some hypothetical idealistic scenario in the future – is that AI is being cited as the main factor behind job losses, redundancies and restructures.

In 2025, for example, while there was an uptick in the number of vacancies for senior or specialized engineers, entry-level and junior roles were decimated, with Crunchbase tracking the number of roles that disappeared.

Many organizations today share the same quiet paradox: the systems aren’t broken, but the business is increasingly operating around them instead of through them.

On spreadsheets. In chat threads. Inside workarounds that started as temporary and quietly became permanent.

The more this happens, the system of record becomes a system of reference, and the real operating system of the company lives in the gaps between the tools you bought to run it.

And this is far from a system of execution.

But this isn't an indictment of the implementation itself. The problem is the clock, and most systems weren't built to keep pace with it.

Many enterprise systems are still designed and deployed on a cycle that assumes business conditions hold roughly steady from the day the requirements are written to the day the system goes live.

But the pace has changed. Customer expectations evolve regularly. New channels appear. Competitors ship in weeks what used to take much longer.

By the time an implementation reaches production, the business priorities it was scoped against have already shifted, and the gap between what the system does and what the business now needs starts widening from day one.

Problems of complexity

This pattern holds across different organization sizes. Large enterprises run into complexity. Mid-sized and smaller organizations run into constraint. Teams implement what the budget or resourcing allows, knowing out of the gate that the system won't be able to cover every need, and they aren’t able to come back to address the rest in a timely or effective manner.

The workarounds start earlier and proliferate faster, but the destination is the same. The system becomes a partial reflection of how the work actually happens, and the rest of the work moves into whatever tool people can get their hands on.

The same drift shows up in how customers experience the company. Too many enterprise systems were designed from the inside out, built around how the organization structures itself internally. Customers don't care about that. They care about being able to buy, get service, or change their account in the way that makes sense to them.

When the system can't adapt to that, the friction shows up in places the company can see clearly: longer call times, dropped journeys, channels that don't talk to each other, customers who have to repeat themselves. The gap between how the business runs and how the customer wants to engage becomes the most visible failure point.

And as organizations expand their AI adoption, the picture gets even more complex. Every agent you deploy will surface a workflow that needs adjusting. Every adjustment will spawn three more. The teams running serious agentic AI pilots are discovering that the AI itself is not the hardest part.

Reshaping the underlying process at the speed the AI demands is what stalls or breaks them. A system whose change cycle is measured in quarters cannot keep pace with AI-driven operations that evolve in days.

How AI upsets the economics

AI is also upending the traditional unit economics of execution. For decades, enterprise systems scaled by adding more users, and more employees meant more capacity. But organizations are now entering a model where humans and AI agents operate together inside the same workflows and operational processes.

As a result, scale is no longer determined primarily by workforce size or application access. It is determined by how effectively the enterprise coordinates execution across people, agents, systems, and automation in real time.

This shift changes the role of the system itself. Instead of serving primarily as a system of record or control, it has to become the place where execution takes place. This moves the focus towards enabling outcomes and orchestrating end-to-end processes rather than simply managing access and data. This is a fundamental shift in how organizations think about both technology and scale.

Faced with these challenges, traditional playbooks for future-proofing now feel woefully inadequate. The five-year roadmap, the multi-year transformation program, the carefully sequenced rollout designed to deliver stability years from now. These were sensible strategies in an era when conditions changed at a pace where you could plan for them. But the pace of the AI era has removed that luxury.

A different operational model

What this moment demands is a different operational model altogether: the Unlimited Enterprise. A platform flexible enough to let teams quickly adjust workflows, introduce new capabilities, and respond to change without long delays.

An environment with no fixed ceiling on users, agents, workflows, or scale, with governance and observability built in. Starting doesn't mean replacing everything out of the gate. Look for places where the friction is greatest, manual effort is heaviest, or readiness is highest, and hit those first.

The companies that win in this next era are the ones engineering the ability to evolve their systems and workflows at the speed business is already moving. Systems that can evolve more easily are more likely to be adopted, remain relevant, and support the business effectively over time. That is the only real future-proofing left.

We list the best small business software.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Artificial intelligence has officially entered its execution phase. After years of experimentation, businesses are rapidly deploying AI not just to analyze data, but to act on it.

At the forefront of this shift are AI agents, autonomous systems designed to execute complex tasks, automate workflows, and interact with other digital systems on our behalf.

Their adoption is accelerating at an incredible pace, with a recent McKinsey study finding that 62% of organizations are already experimenting with them. It’s easy to see why, as Agentic AI offers a relatively straightforward path to embedding powerful automation deep into business processes.

However, as these systems evolve from passive tools into autonomous agents, we are entering a new era of digital risk. The conversation is no longer just about building smarter agents, but whether the entire agent economy can function without a trust layer underneath it. Identity verification has evolved beyond being a basic security feature into the infrastructure that makes autonomous commerce possible at all.

When an AI can access sensitive databases, interact with third-party systems, and execute commands, a critical question emerges: how do you know who, or what, is really on the other end of that API call? Without a robust framework for agent identity and accountability, true agentic commerce will not be possible.

The real-world risks of unchecked AI agents

The appeal of agentic AI is its autonomy, but this is also its greatest risk. An unchecked or compromised AI agent operating within a corporate network can become a powerful vector for malicious activity. These risks are not new, but are a direct extension of existing cyber threats, amplified by the speed and scale of AI.

The most defining threat vector in the era of agentic AI is prompt injection. This is unique to AI agents as it weaponizes the very natural language capabilities that make them so powerful. Unlike traditional cyberattacks that rely on exploiting software bugs or cracking passwords, prompt injection bypasses standard security perimeters by feeding maliciously crafted text directly into an agent's processing stream.

This essentially tricks the AI into overriding its core system instructions and executing the attacker’s commands as if they were legitimate tasks. In an enterprise environment where agents hold permissions to access CRMs, process invoices, or alter databases, a successful injection can instantly turn a helpful digital assistant into an undetected insider threat.

Through prompt injection, agents can be instructed to exfiltrate data or escalate their privileges. An agent designed to access a customer database for legitimate analysis could, if compromised, be instructed to copy and transmit that entire database to an external server. Similarly, privilege escalation becomes a major concern, as an agent with limited permissions could probe the network for vulnerabilities or exploit a flaw to grant itself higher levels of access, effectively becoming a rogue administrator.

AI-to-AI interactions present a new frontier of security risk. As one business’s AI agents begin to interact with agents from partners or customers, the potential for supply-chain compromise grows exponentially. Without a way to verify the identity of the interacting agent, every AI-to-AI connection becomes a potential security blind spot.

Building a framework for agentic trust

In the emerging agent economy, trust hinges on answering three questions, only two of which today's standards meaningfully address: “who is this agent?” (addressed by identity primitives like W3C DIDs, increasingly applied to agents), “does this agent have authorization to spend this money on a user's behalf?” (frameworks such as FIDO Alliance-stewarded standards such as AP2 and Verifiable Intent, contributed by Google and Mastercard), and finally “what is this agent's reputation and track record?”, a question the current standards stack leaves open.

Together, they form the essential trust and payment stack required to move agentic commerce from experimental sandboxes to mainstream, high-value transactions. Zero trust architecture is also more critical than ever for securing systems against agentic threats. An agent's identity must be re-verified for every single transaction or request, and its permissions should be limited to the absolute minimum required for its specific task, based on the principle of least privilege. This means even if a trusted agent is compromised, its ability to cause widespread damage is severely restricted.

This same logic extends well beyond the corporate perimeter, in both directions. On one side, AI agents are beginning to transact on behalf of consumers: booking, buying, paying, returning. On the other, businesses are deploying agents to fulfil those same orders, onboard new customers, automate supply chains, and run entire back-office functions. What's emerging is a new trust triangle between consumers, businesses, and agents, operating simultaneously on both sides of every interaction.

In that world, agent identity becomes a commercial problem as much as a security one. A business needs to know that the agent placing an order holds a valid, scoped mandate from a real human who authorized it to act. But equally, a consumer's agent needs confidence that the business agent fulfilling that order is legitimate, authorized, and traceable. Trust has to flow in both directions, and at machine speed. That's a verification challenge of a fundamentally different order to anything we've dealt with before, and one the industry is only beginning to standardize through frameworks like FIDO's Agentic Payments Protocol. Getting KYA right is foundational to enabling a function agent economy.

Finally, businesses need systems that continuously monitor agent behavior to create a baseline of normal activity, making it possible to spot anomalous actions. If an agent suddenly attempts something outside its regular function, such as accessing a new database, connecting to an unusual IP address, or executing commands at a much higher frequency, this behavior should instantly trigger an alert and, potentially, an automatic suspension of the agent’s permissions.

Trust as a catalyst for innovation

Some technology leaders hold the view that strict security measures are a barrier to innovation, however in reality the opposite is true. By building trust and safeguards into AI agents from the ground up, businesses can innovate without fear. They can confidently deploy agentic solutions to drive efficiency, reduce operational costs, and unlock new revenue streams, all without exposing themselves to the catastrophic risks of uncontrolled autonomy.

The agentic AI era is here, and it has the potential to reshape how enterprises operate. Autonomy without oversight is liability, but autonomy with verified identity, scoped mandates, and continuous trust signals is the foundation of a new commercial layer. As agent architectures mature, trust certification will become a precondition for being transacted with at all. KYA isn't a security cost. It's how you stay in the game.

We feature the best ecommerce platforms.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

“With great power comes great responsibility.”

It may be a line borrowed from Spider-Man, but it captures the inflection point facing the telecoms industry today.

Telecom networks are the invisible infrastructure of modern life. Every message sent, video streamed and connected device depends on the vast digital highways that telecom operators build and maintain.

But the industry is now at a crossroads.

As UK telecoms intensify their journey toward Net Zero and with 6G on the horizon, AI is emerging as both a transformative enabler and a sustainability challenge.

A dual approach to AI

Telecoms has entered a new phase of AI-fueled growth. Data demand is surging, networks are becoming denser, and the next wave of innovation is pushing digital infrastructure to new limits.

With this growth comes a clear opportunity to lead on sustainability – but to seize it, the industry must unlock AI’s full potential responsibly. On one hand, AI enables smarter operations across the telecoms system, from network traffic optimization to supply chain analytics and carbon reporting. On the other, it brings rising energy and water demands driven by compute-intensive workloads.

A dual approach is therefore central to ensuring technological innovation supports the climate agenda. That means advancing both AI for sustainability – using AI to reduce emissions, optimize resources and accelerate progress towards Net Zero – and sustainable AI, ensuring that AI itself is designed, deployed and governed in a way that minimizes its environmental impact.

This is the defining tension of the next decade: scaling intelligence without scaling impact.

Pinpointing high-impact AI use cases

AI is already creating new possibilities for progress, but real value lies in prioritizing use cases that drive both business value and environmental benefit. Not every use case delivers equal value—and in a resource-constrained world, prioritization is critical.

First and foremost, energy optimization is an area of significant opportunity. Energy consumption accounts for up to 40% of telecoms’ network OPEX, and GSMA Intelligence indicates that operators could reduce OPEX by 4% for a 20% reduction in energy costs through power efficiencies.

Here, AI can assist with network load balancing to help distribute traffic evenly and reduce unnecessary energy consumption. Meanwhile, energy use in data centers can be reduced through intelligent workload distribution, predictive cooling and server optimization, all of which lower overall demand. In effect, AI enables networks to think more intelligently about how and when energy is used.

On the reporting side, AI is also playing an increasingly important role. It’s being used to support customer carbon reporting, particularly in analysing large datasets from multiple sources – estimating notoriously obscure Scope 3 emissions and helping customers better understand their impact.

For example, a network operator seeking to assess B2B customer emissions can integrate data from traffic records and device energy use, linking it with customer profiles to generate more accurate insights. What was once opaque is becoming measurable, and therefore manageable.

Establishing the foundations for sustainable AI

Without the right foundations, scaling AI risks putting the cart before the horse. Organizations must establish strong governance to ensure AI is used ethically and sustainably – and that requires close collaboration across sustainability, technology and operational teams.

A centralized governance body can accelerate decision-making, enforce accountability, and ensure AI initiatives are aligned with wider sustainability goals. It also provides the guardrails needed to scale AI with confidence.

This oversight should extend across the full AI lifecycle — from design and development to training and deployment — embedding environmental, social and ethical considerations from the outset, rather than treating them as an afterthought.

For instance, organizations can adopt a “decision tree” approach to evaluate whether AI is needed at all, and if so, what type is most appropriate, considering alternatives that may deliver similar outcomes with a lower environmental footprint.

Assessing and mitigating environmental impact

Understanding AI’s environmental footprint must be the starting point for responsible adoption. After all, you can’t manage what you can’t measure. In truth, only a minority of companies are actively monitoring and disclosing the environmental footprint of their AI models, and even fewer have set reduction targets. This indicates a broader need for lifecycle assessments and accountability in AI operations.

But this is a daunting task on a few levels. Many organizations lack transparency from AI providers regarding energy efficiency and carbon footprints, which is crucial for informed decision-making – and without clear, standardized metrics and greater visibility across the value chain, organizations are left making critical decisions in the dark. Addressing this gap will require closer collaboration across the ecosystem to establish common standards, improve transparency and embed sustainability into AI decision-making.

A responsibility that extends beyond telecoms

The impact of telecoms sustainability decisions goes far beyond the industry. Telecoms already account for at least 1.6% of global carbon emissions. Decisions made here influence energy demand, emissions reduction and resource efficiency across national economies. By acting decisively, telecoms can set a standard for other sectors, demonstrating that AI-powered growth and sustainability are not mutually exclusive.

But there is work to be done. As AI adoption accelerates across networks, customer operations and enterprise services, understanding and managing its energy impact will become increasingly important. Ultimately, the challenge facing telecoms is not simply to connect the world, but to do so responsibly.

We feature the best business phone systems.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

I’ve always favored clean, minimal phone homescreens — the fewer icons, the better, with the wallpaper fully visible. This usually means I have to delete a lot of icons and widgets after initial setup, especially with a Samsung Phone and an iPhone to achieve that look. Pixel phones are much less cluttered in comparison, though I still have to tweak some settings a bit to personalize the homescreen fully.

Then, a few months back, I got my hands on the Nothing Phone 4a Pro, and it was love at first sight. Not only does it look great and punch above its weight class with its hardware, but Nothing OS seemed tailor-made for me — it’s now my favourite Android skin.

I found the monochrome color scheme and minimal interface very appealing, reminding me somewhat of Pixel UI’s stock Android experience but with more character. The home screen widgets follow the same design philosophy, and community-made ones add extra functionality and whimsy I didn’t know I needed.

Clean, distraction-free bliss

My usual homescreen setup features just a single row of my most-used apps at the bottom, with icons set to a monochrome color scheme and maybe one or two widgets.

These settings are readily available on iOS and Pixel UI, and easy to apply, while Samsung’s One UI has a wide range of icon packs to choose from. But where you have to manually opt to have this minimal look, it’s the default on Nothing OS. It’s in no way a groundbreaking innovation, but it’s my favorite part of using the Nothing Phone 4a Pro.

Nothing OS widgets also set themselves apart from typical Android or iOS widgets — they come in small square tiles (just enough to fit 4 icons in a 2x2 arrangement) that match the icons’ look. Even when I load up the 4a Pro’s homescreen to the brim with widgets and icons, it still looks clean and slick.

Don’t care for the default widgets? No problem. There are plenty of community-made ones you can download from Nothing Playground, a portal for widgets (found under Essential Apps in Settings) and equalizer profiles for audio. These are unique and can show off your personality too. Some of my favorite community-made widgets include a tic-tac-toe game, a functioning piano, and a meter that tracks CPU usage and temperature.

The latest major Nothing OS update introduced simple breathing exercise widgets — one each for Focus, Calm and Relax — with on-screen prompts that tell you when to inhale and exhale while music plays. Again, it’s not revolutionary, but I like being able to tap a widget on the homescreen instead of digging through menus in a dedicated health app.

These little things add up to making a fabulous user experience in my books.

Innovations and experiments

In my opinion, an even more impressive addition in that update is Essential Voice, an AI-powered speech-to-text feature that has produced the most accurate results I’ve seen on a phone so far.

Powered by Gemini 3 Flash, Essential Voice automatically removes filler words and sounds like “um”, “ah” and “basically” — something I tend to do a lot myself — and also auto-formats bullet lists. I tested this feature by dictating my rather long, rambling notes into Google Docs and it produced a neat outline to build the article I was working on.

There’s support for more than 100 languages and regional variants, which I tested by reciting phrases in Spanish, Italian, French and Filipino — the results were clean and accurate. While a similar feature called Rambler is coming to more phones via Android 17, it’s nice to see it already in action on Nothing OS. While I’m not quite ready to forgo keyboards just yet and talk to my phone instead, I still think it’s a fantastic shortcut to typing.

Nothing OS also lets you try experimental features like using the Glyph Matrix (the secondary screen on the 4a Pro’s back) as a progress bar for third-party apps (it only supports Uber, Zomato and Google Calendar for now) and improving Apple AirPods support — well, for an Android phone at least. They’re not the most earth-shattering features, sure, but they’re a promising sign of Nothing’s willingness to try new ideas and let users play around with them.

Admittedly, Nothing OS’s quirky design won’t be for everyone, as some users will prefer more conventionally designed widgets or colorful icons that are easy to identify, but the short 3-year software support window is what’s most disappointing here, especially when Samsung and Google offer 7 years, while Apple is pushing that to 8 years in some cases.

Nevertheless, Nothing’s phones stand out as aesthetically pleasing handsets that punch above their price range, with the software experience as the unheralded star of the show — at least for me. It’s the perfect minimalist experience out of the box, without sacrificing functionality, and it includes some useful extras enhancing the experience.

With Nothing changing its flagship release schedule for 2027, I’m curious to see how Nothing OS will evolve alongside a potentially more powerful phone. Perhaps it will offer more ways to encourage community innovation? Either way, I’m certainly keeping my eyes peeled for what’s next.

The iPhone will go down in history as one of the most iconic technologies that any business has introduced to the world. But not everybody, including Microsoft's former CEO Steve Ballmer, thought it would be a hit.

Disrupting the market

Steve Jobs' iconic presentation revealing the iPhone in January 2007 had an air of magic around it, capturing the attention of plenty of commentators and technologists.

But there were also plenty queuing up to label it a fad that would never take off. Steve Ballmer, Microsoft's CEO at the time, was among them.

In an interview with USA Today, a grinning Ballmer laid into the new device, criticizing its expensive pricing and the lack of a physical keyboard. This is a feature that Microsoft supported with its Windows Mobile operating system that it licensed to third-party developers like Motorola, which itself was selling a $99 Motorola Q – something that Ballmer considered a much better business alternative.

The iPhone marches on

To be fair to Ballmer, he did say in the same interview that the iPhone could end up selling really well. But he certainly failed to catch onto its appeal not only as a popular device but one that businesses would embrace.

Businesses appreciate the intuitive mobile device management (MDM) system backing iPhone in the workplace and the interoperability between Apple devices is something many also find useful. That's not to say Android devices aren't suitable for businesses — but Apple indeed tops the rankings for the highest number of shipments among enterprises.

Ballmer has since conceded that he made the wrong call – and it just goes to show how tricky predicting the future could be.

We now know with more certainty than before that your next iPhone, especially if it's a more affordable model, will probably be more expensive come September.

Allow me to walk you back through my reasoning.

At approximately 8:30AM ET (that's 10:30PM AEST) the RAM crisis reached Apple shores in the form of price hikes across multiple product categories, including MacBooks, Macs, iPads, and HomePods.

It was inevitable and, as Apple told us, the culprit is clear: "The rapid expansion of AI data centers has created an extraordinary surge in demand for memory and storage."

It's the component crunch we've seen repeated over and over again. Everything from the latest gaming rigs to storage and memory is getting more expensive.

And it's only going to get worse.

"We are not at the bottom and will take more time to climb out," wrote Creative Strategies Founder and longtime analyst Tim Bajarin when I asked him via email if this marked a tipping point for our RAM crisis travails. It did feel like Apple held off as long as possible, and I think Bajarin concurs, "Apple had no choice," he wrote to me.

No RAM crisis end in sight

Not only does Bajarin see the issue continuing, but he thinks it could drag on for years. As he wrote in his recent Forbes column, with just a few major memory factories already "maxed out" and those under construction years away from coming online, "I see this memory squeeze at the least lasting another two years," he told me via email.

This aligns, by the way, with reports we've seen elsewhere from those who run these memory plants.

"We expect tight conditions to persist beyond calendar 2027 as a result of AI-driven demand across all segments coupled with structural supply constraints," said Micron CEO Sanjay Mehrotra in a recent earnings report.

We've all been wondering if and when the RAM crisis would affect iPhone prices, and while the fresh price hikes miss out iPhones, Apple Watches, and AirPods, Apple's comment on why it raised the prices now, makes it clear that they're probablly not done: "We have shielded our customers from these increases so far, but we have now reached a point where we need to begin raising prices on a number of products, including today’s increases for iPad and Mac."

Did you pick up the key phrase? "need to begin raising prices". Apple didn't say, "we've raised the prices," and instead opened the door to further hikes.

Those price increases might come to the full lineups in affected product categories, but I believe that this issue will affect the anticipated iPhone 18 launch in September.

But wait. It gets worse.

An affordability crisis

Bajarin says the price/performance trajectory we've seen for decades (more and faster/better tech for lower prices) is, with this component crisis reversing itself, making it harder to build affordable devices, or rather devices at affordable prices, because one of the key components is exponentially more expensive than ever. We already have some evidence of this with Nothing cancelling its affordable model in response to RAM crisis.

This means that those who usually buy the iPhone Pro, iPhone Pro Max, and iPhone Air might easily absorb higher prices because they can probably already afford them.

It's with base models like the anticipated iPhone 18 where a dramatic increase could push the phone out of reach. This will, by the way, affect Android phones too (see Nothing, above), especially, Bajarin notes, those affordable handsets sold around the world.

You see, they all source their memory from the same handful of suppliers. It's all in shorter supply and more expensive.

In the end, if any of us thought Apple could hold out indefinitely and show the industry a way around this AI-led RAM crisis (the AI boom isn't slowing down, and its demands on energy, water, and memory will probably only increase), they were mistaken. Apple isn't magic, RAM doesn't grow on trees, and we're in for a very rough couple of years.

A look at all the Apple price hikes

Logistics operators are under pressure from every direction.

Warehouses are being asked to process higher volumes, respond faster to disruption and maintain tighter inventory control, all while managing labor shortages and rising operational costs.

UK Warehousing Association research shows that recruitment challenges continue to affect the sector, with only 13% of employers reporting no difficulty hiring staff and more than half anticipating critical skills shortages in the years ahead.

At the same time, investment in warehouse automation continues to accelerate. McKinsey estimates adoption is growing at more than 10% annually as operators look to improve efficiency, resilience and cost management across increasingly complex supply chains.

Yet the defining shift in logistics is no longer the rollout of isolated automation tools. It is the emergence of autonomous systems capable of continuously capturing operational data, generating live insight and supporting faster, more informed decisions across the warehouse environment.

Moving beyond static automation

Historically, automation in logistics has been limited to controlled and predictable settings. Fixed systems deliver value in stable workflows.

That is beginning to change.

Most warehouse environments are dynamic, with changing inventory, layouts and operational demands. Advances in artificial intelligence, sensing and mobile robotics are enabling more effective operation in complex environments. At the same time, digital modelling technologies are improving how operators understand and manage warehouse operations.

Gartner predicts that by 2030, half of new warehouses in developed markets will be designed as human-optional facilities, supported by robotics and digital twins. This does not signal the removal of people from logistics. It reflects a shift towards systems that can support human decision-making with more accurate and timely data.

Creating a live operational picture

One of the biggest constraints in warehouse operations is visibility. Many decisions still rely on periodic stock checks, incomplete system data or manual investigation.

Digital twins are beginning to address this challenge.

By combining continuous data capture with virtual models of the warehouse, digital twins provide a real-time view of inventory, storage and movement. This allows operators to detect discrepancies earlier and respond before issues escalate.

Deloitte research indicates that improved inventory visibility can reduce operational inefficiencies and improve fulfilment accuracy, especially in high-volume distribution settings. More broadly, the ability to maintain accurate, real-time inventory data is directly linked to warehouse performance. Industry studies predict that poor inventory accuracy can significantly increase costs through mis-picks, stockouts and excess safety stock.

Turning data into operational action

The value of autonomy increases when insight leads directly to action.

As real-time data is combined with analytics, warehouse systems can identify issues such as misplaced inventory, congestion or underutilized space as they occur. This enables faster intervention, whether through human decision-making or system-driven recommendations.

Over time, this creates a more responsive operating model. Instead of reacting to problems after they occur, warehouses can move towards continuous optimization.

McKinsey notes that advanced analytics and automation can significantly improve warehouse productivity and reduce operational costs when deployed effectively.

Scaling logistics intelligently

A persistent challenge in logistics is that growth often brings added complexity. Higher volumes typically require more labor, more oversight and more manual processes.

Autonomous systems offer a way to change that dynamic.

By automating data capture and improving decision-making, organizations can scale operations without a proportional increase in manual intervention. This is particularly important in a constrained labor market.

At the same time, newer forms of automation are designed to be more flexible than traditional fixed systems.

Autonomous mobile robots, for example, can operate within existing warehouse environments, eliminating the need for large-scale infrastructure changes and making adoption accessible more easily.

How autonomy is changing warehouse roles

Autonomy does not replace people. It changes how they contribute.

As repetitive and physically demanding tasks are automated, workers can focus more on oversight, exception management and continuous improvement. This shift is already visible in facilities where humans and machines operate together, combining physical automation with human judgement.

Safety is also a key factor. Warehousing remains a high-risk environment, with industry data showing that the transportation and storage sector, which includes warehousing and distribution, experiences above-average non-fatal injury rates and around 38,000 workplace injuries each year. Reducing manual handling and improving situational awareness through automation can help mitigate these risks.

Building resilience through visibility

Autonomy is becoming a foundational capability for modern logistics operations. The greatest value does not come from robotics alone, but from the ability to combine automation, AI and real-time operational intelligence into a single connected system.

Warehouses that can continuously monitor inventory, identify inefficiencies and respond dynamically to changing conditions will be better positioned to manage disruption and scale sustainably.

As supply chains become more complex and less predictable, the competitive advantage will increasingly belong to organizations that can make faster and more accurate decisions. The future of warehousing will depend on how effectively businesses turn operational data into meaningful action.

We've listed the best ERP software.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Banks aren’t short on AI ambition. Across the industry, there are almost weekly announcements about new deployments, new partnerships, and new capabilities.

But amidst the lofty pronouncements, not enough attention is given to building the underlying architecture that makes any of it sustainable.

The UK Treasury Committee has already warned that the financial system is not adequately prepared for a major AI-related incident, and from where I sit, that warning is aimed squarely at how modernization programs are being run.

This is a governance problem which stems from unmade decisions; most banks have yet to make the strategic, organisation-wide commitment that should determine what gets built, before anyone starts building.

There is a real opportunity to be grasped by inviting the risk team to be part of design before you begin delivery.

Risk teams are being invited to the table too late

In traditional program structures, risk and compliance are brought in to validate decisions already made. Platforms are selected, designs agreed and timelines set, and the role of the risk function has typically been to check these things, rather than shape them. That model was never ideal, but now, in the AI-era modernization push, is structurally backwards at best and dangerous at worst.

A genuinely AI-ready bank requires continuous data lineage, runtime-embedded controls and a real-time, cloud-native core. None of these can be retrofitted cheaply once architecture decisions have already been made. By the time a compliance team discovers that a new platform can't meet AI governance or data traceability requirements, the cost of fixing it is prohibitive

If the risk function is part of the design process from the outset, banks can build a risk-conscious architecture rather than a risk-adjacent one. The difference in outcome is substantial, and increasingly, the difference between an institution that can operate confidently with AI and one that can't.

How legacy cores create invisible risk

The story of legacy architecture's constraints on innovation is well told, but it also creates blind spots that risk professionals can't see, let alone manage.

Batch-processing cores are the clearest example of this. When a core updates positions overnight rather than in real time, AML and fraud systems are operating on stale data. Suspicious activity that occurs between batch runs is invisible until the next cycle, a direct operational liability as the volume and speed of financial crime increases.

Fragmented data pipelines create a related problem. Continuous lineage (the ability to trace every decision to its data source) is a prerequisite for AI governance. On architectures where data moves through multiple disconnected systems before reaching an analytics layer, that lineage is structurally impossible to maintain without a modern, cloud-native architecture.

Finally, third-party AI deployment adds an additional layer of exposure. When models are embedded in platforms a bank doesn't have full control over, it leaves compliance teams operating with limited visibility into how decisions are being made.

Under the EU AI Act, which requires traceability and explainability for high-risk AI systems which may credit decisioning and fraud detection, that becomes a direct regulatory risk, and that's just one of a raft of emerging AI regime banks must navigate.

Risk is moving from chronic to acute

What has changed is the speed at which risk moves.

Banks now have to contend with an increase in state-based cyberattacks on financial supply chains, a trend accelerated by geopolitical instability, with agentic AI substantially increasing the speed and scale of those attacks.

Under regimes such as DORA, which introduce near real-time cyber incident reporting, means institutions need detection and response infrastructure that operates on live data, capable of identifying and escalating incidents as they happen rather than after the fact. Batch architectures don't meet this standard, and the window to remediate is narrowing.

The EU AI Act raises the bar further on traceability and explainability, particularly for credit decisioning and fraud detection, both areas where AI adoption is moving fastest.

Banks not designing against these requirements are accumulating risk, not avoiding it. The regulatory deadlines are known. The technical requirements are understood. The question is whether the architecture being selected today is capable of meeting the obligations that will be enforced tomorrow.

What risk-intelligent modernization looks like

I have sat in enough program steering committees to know that the moment risk professionals are brought into a modernization program shapes everything that follows.

When we arrive after the architecture is set, we are negotiating against decisions that are already expensive to reverse. When we are in the room from the start, we can design the non-negotiables in rather than bolt them on.

The practical starting point is three questions that every bank should answer before any platform decision is made:

What risk posture must we preserve throughout migration? The answer defines the non-negotiables for any new architecture: data integrity, audit continuity, access controls, and the governance structures that cannot lapse during transition.

What are we engineering against on a ten-year horizon? Regulatory requirements in 2036 will look different from those in 2026. The architecture being selected today needs to be capable of meeting obligations that do not yet fully exist, which means flexibility and real-time data capability are not optional extras. Some of the newer, digitally-native fintechs have demonstrated that building with this horizon in mind from the outset produces materially different, and more resilient, architecture than retrofitting compliance onto an existing estate.

What governance must change so we don't embed legacy risk into a modern platform? New infrastructure running old processes is window-dressing. If you modernize technology without modernizing governance, you recreate the same control failures in a different environment.

Modernization is not risk-free. But the risk of standing still now exceeds the risk of moving carefully, and the cost of getting the architecture wrong compounds with every year it goes unfixed.

Store you business data in the cloud with the best business cloud storage services.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

The potential around AI in finance is tangible right now and with it comes both opportunity and risk.

Vendors are promising faster closes, automated reporting and real-time insights. Finance leaders are under pressure to modernize and keep pace with the AI revolution.

Across the mid-market, AI is being embedded into financial workflows at a pace that is outrunning the governance structures needed to make it safe.

That gap is where a new category of risk is emerging, one that for professionals working closely with finance teams are increasingly calling "black box AI." It is yet another challenge the industry needs to address head-on.

Black box AI refers to any finance deployment where outputs cannot be traced or governed. It shows up as unverifiable recommendations, automation that bypasses normal approval workflows, or analytics tools that surface a headline figure with no clear path back to the underlying transaction.

The outputs look credible on the surface, but when an auditor asks what has changed, who approved it and why, that’s when finance teams run into problems.

Finance teams have always operated on a simple principle: if you cannot trace it, you cannot trust it, this mantra has not changed in the age of AI. What has changed is that the technology can now either reinforce that discipline or undermine it.

Which outcome occurs comes down to how AI is built and what your provider is willing to be held accountable for.

Assistive versus autonomous: A distinction that matters

Not all AI carries the same risk profile. The distinction for finance leaders is between AI that assists in a controlled environment and AI that acts autonomously.

Assistive AI flags issues and directs decisions through existing approval structures. It identifies an exception, spots a coding anomaly, or highlights an unusual movement, and then puts a human in the driving seat for decision-making. The finance professional can see the source transaction, the rule behind the flag, and the approval record. Every action the AI makes also leaves a clear audit trail.

Autonomous AI, by contrast, makes decisions silently. It allocates, codes, or adjusts without a visible audit trail, leaving teams unable to satisfy even basic reporting requirements. It may make things look faster but it is also quietly undermining the integrity of the numbers.

The question every finance leader should be asking their technology provider is simple: if something changes in our reporting, can we see exactly what happened, who approved it and why? If the answer is unclear or non-existent, then that is a problem for the finance team to address.

Where governed AI delivers

Month-end close is where well-governed AI delivers the clearest practical return on investment, and it is worth being precise about what that looks like.

The real win is not AI-generated narratives or automated reporting. It is catching issues earlier in the process before the stakeholder board packs go out, before the pressure peaks and before the rework begins. When AI surfaces exceptions at the right moment and directs them through channels, teams can close faster and with greater confidence in what they are signing off.

For organizations operating across multiple entities or currencies, this becomes even more valuable. AI can flag unusual movements at subsidiary level before they surface as problems at group consolidation but only when the underlying model is governed properly and the exception trail is intact.

Raising the bar on what ‘finance-grade’ means

There is a version of AI in finance that makes teams more effective, then there is a version that makes things look faster while quietly compromising the controls that finance exists to uphold.

The industry needs to be more precise about what earns the label ‘finance-grade AI.’ Reporting or analytics tools only deserves that description when a professional can analyze from a headline variance figure down to the underlying transactions and from there to the audit-trail evidence that can explain that headline variance number.

Anything that breaks that process - however sophisticated it appears - is not fit for purpose in a governed finance function.

Finance leaders cannot afford to adopt technology simply because it is what everyone else is doing. The right question is not whether AI can automate a process, but whether it can do so in a way that keeps the finance team genuinely in control of their numbers with a ‘human in the loop’.

That standard should be a non-negotiable and it should be first on the checklist before any AI deployment goes live.

Protect your computer or network with the best internet security suite.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

The semiconductor industry has two competing priorities.

On the one hand, demand is soaring—last year’s sales were up 25.6%. 2026’s sales should cross the $1 trillion USD mark. Meeting this demand means increasing production—and production is a resource-intensive process. Direct emissions increased between 2023 and 2024.

Projections indicate that by the critical climate year of 2030, manufacturing chips will come with emissions totals larger than those of some countries.

It seems simply impossible to both reduce emissions while increasing the volume of chip production.

But efficiency-minded manufacturers are proving it’s possible to do both.

To solve this energy paradox, the industry needs a holistic four-step process: measure, replace, reduce, and engage.

1. Measure

While any manufacturer around the world can go through this process, every fab is different. There’s simply no one-size-fits-all solution to reducing emissions at every facility around the world.

Adequate measuring provides meaningful, context-sensitive data to guide decision-making. While it’s an older acronym, WAGES (Water, Air, Gas, Electricity, Steam) is a good foundation for what needs to be measured. Life Cycle Assessments (LCAs) for specific products like a die or wafer are also useful to have.

Complete, high-quality data is the key to several key steps for effectively reducing emissions. First, it enables comparing product lines and facilities, which in turn helps create realistic KPIs for emissions reductions.

Second, any efforts to optimize energy usage via AI will fail without sufficient data. Third, the right kind of data enables third-party confirmation and subsequent credibility, such as obtaining validation from the Science Based Targets initiative (SBTi).

2. Replace

As imec has reported, manufacturing chips at scale requires an enormous amount of electricity. Given that only about a third of global electricity comes from renewables, this means electricity consumption is also a major contributor to facility emissions.

Yet it’s possible to improve the energy supply mix. In one specific example, Nvidia earlier this year confirmed 100% renewable energy usage at its controlled facilities. But even companies that aren’t operating at Nvidia’s scale (or almost every company on Earth) can make choices that produce meaningful impact.

Negotiating a Power Purchase Agreement (PPA) during the design phase can be a critical step to ensuring long-term renewable power supply—while also insulating your facility from future price shocks. If possible, band together with other smaller companies to increase your negotiating power.

However, replacing equipment (or choosing more energy-efficient equipment to start) is another way to quickly cut energy consumption. For example, a traditional chilled water cooling system can be a major power consumer. But McKinsey identifies several potential alternatives or upgrades: adding in idle-time controllers, using ambient air and cooling towers, and meters to identify potential areas for improvements.

These two tactics address the issue both before power even reaches the fab and during manufacturing—and can save both capital and emissions.

3. Reduce

Adding meters and controllers also simply reduces the amount of energy used in the first place. By taking the time to collect, measure and organize data, fabs can precisely determine the best targets for reduced usage and lower emissions.

In conjunction with hardware like meters and well-organized data, AI can be a helpful tool. With the right data, it can analyze power usage patterns, including unusual spikes, and generate recommendations. It can also enable predictive maintenance, which can reduce costs and emissions in the long run by identifying problems that can be fixed before they get too resource-intensive and extending equipment lifecycles.

Precise targeting also enables precise measuring for long-term success. Key net-zero target years like 2030 (Paris Agreement) and 2040 (the Climate Pledge year) are still relatively far away at the speed of technological innovation. Without baseline data or an action plan, it’s difficult to ensure that goals are actually being met—and a lack of concrete evidence could compromise credibility with investors, customers and employees.

4. Engage

Reducing emissions is more effective when it’s done in collaboration. Working closely with partners, suppliers, different site managers and other stakeholders can surface new ideas and solutions.

For example, one potentially overlooked area to target is the supply chain. Chip components often take multiple trips on a plane for different operations.

Employees can also be an excellent source of information. With on-the-ground experience, they may have additional insights into how operations play out and what can be done to improve them.

Engagement isn’t a one-and-done process, either. Plan to regularly communicate results to employees, partners and customers. This can not only lead to stronger relationships but also improved employee retention and even public recognition.

When it comes to reducing the semiconductor industry’s emissions, there is no single solution. The current level and projected increases are simply too large to focus only on one area, like energy sourcing, equipment or logistics.

Instead, reducing fab emissions must be a multi-step, multi-faceted journey. But with the right measurements in hand, fabs can precisely maximize the impact of every replacement, reduction and engagement.

The best cloud storage - our expert rankings.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Operational performance is becoming just as important as investment performance in private credit.

As fundraising slows and investor expectations increase, firms are facing growing pressure to modernize the IT infrastructure supporting their portfolios.

In fact, transparency and faster reporting are becoming top priorities.

Without such capabilities, funds can’t clearly see across their own portfolio, amplifying stress when markets are less forgiving.

Especially as private credit scales and operational diligence become more central to allocation decisions, such back-office issues become something more structural.

Fortunately, managers that properly invest in their operational foundations now, will be better equipped to manage the increasing demands facing the private credit industry moving forward.

Analyzing Operational Pressures

Private credit is operationally intensive. And many firms never built systems to match the increasing complexity of their portfolios as they grew and evolved. Instead, operations are often spread across legacy servicing platforms, spreadsheets, email-driven workflows, and disconnected internal tools, leaving firms without a single real-time view of portfolio data.

Most loan servicing operates on cycles such as monthly reconciliations, quarterly reporting, and batch-based payment processing. This model reflects the constraints of legacy systems and manual workflows. Data must be collected, validated, and processed in stages. Consequently, funds often view their portfolios through periodic snapshots rather than in real-time.

Many funds also maintain parallel spreadsheets to verify their servicer's calculations. Known as shadow booking, this redundancy has no other purpose than to increase control. Ultimately though, it’s a sign of mistrust in the data provided, and the underlying calculations are often difficult to inspect. When discrepancies arise, they are discovered after the fact and require manual investigation, often across multiple systems.

These realities of the private credit industry are now colliding with a more competitive fundraising environment. Even if their underlying deals aren't catastrophic, private credit funds begin to look fragile if they cannot perform operationally well under increasing pressure.

Not surprisingly, many private credit firms are looking to AI to address these issues. But automation software alone cannot repair broken operational foundations. The critical question becomes whether their infrastructure is actually prepared to support the AI model they choose to use.

AI Alone Is Not the Answer

AI agents can execute operational work reliably while platforms can provide real-time visibility and full auditability. The pieces are in place. But there’s a pattern in how AI projects fail in the finance industry that's worth naming and it's almost never the model that's the problem.

What’s often missing is the infrastructure surrounding the model - the systems, workflows, data access, permissions, and controls that allow AI to operate reliably inside real financial processes. Otherwise known as the harness. Generic AI tools don’t know what a rate notice is.

They don't know that a prepayment request triggers a multi-step workflow across every syndicated entity on a facility. They don't know the difference between a funded tranche and a committed-but-undisbursed revolver, or why that distinction matters for how a payoff figure should be calculated.

This lack of operational context is also one of the biggest reasons AI hallucinations occur in finance. The majority of hallucinations aren't random malfunctions, they're a context problem. The AI model wants to give an appropriate answer. When it doesn't have access to the specific data it needs, it reasons from what it does know and produces something plausible.

Private credit portfolio data isn't embedded in any public language model. If the harness doesn't provide it, through tools, memory or real-time data access, the model will fill the gap with something that sounds right. Which, in a financial operations context, is a real problem.

The fix isn’t a better model. It’s a harness that gives the agent access to the right data, at the right moment, with the right tools and controls to retrieve it.

The firms that focus on building operational systems that provide context, transparency, audibility and human oversight will receive the greatest value from its AI investments. In private credit, the long-term advantage may not come from adopting AI faster than competitors, but from building the infrastructure, or the harness, capable of supporting it responsibly and at scale.

Establishing New Competitive Differentiators

As these operational systems mature and advance, they will also reshape what excellence actually looks like inside private credit firms. Responsiveness won't be a differentiator. It will be assumed. Real-time and instant delivery will be the new baseline.

This is because most routine interactions will no longer require human involvement. With real-time systems, shared data layers, and automated workflows, information will be directly accessible and continuously updated. What previously required a request and response cycle will be resolved at the source.

As a result, the role of the servicer shifts. They are no longer measured by how quickly they process or reply, but by how well they handle what cannot be automated - exceptions, edge cases, and judgment calls.

This is why the next generation of private credit leaders may look fundamentally different from the firms that defined the industry’s earlier growth period. Capital access and underwriting expertise will remain essential, but operational execution is becoming increasingly strategic.

Most funds are using general-purpose AI for ad-hoc analysis or are in a holding pattern. A small number are starting to build their own and discovering how much harder it is than first expected. The funds that move first on specialized operational infrastructure will have an advantage that compounds. Not because they picked the right model (the model will keep getting better and cheaper regardless), but because they built or adopted the right harness, trained it on the right context, and gave it the controls that make it trustworthy at scale.

In many ways, private credit firms are evolving into operational organizations as much as financial organizations. The ability to manage workflows, data, oversight, and execution will become a defining part of a firm’s competitive performance.

We feature the best personal finance software.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Data governance is unglamorous work. It is also the reason most AI strategies stall before they scale.

Spending on models, platforms and use cases keeps growing. But the disciplines that make those investments effective – data quality, ownership and governance – often receive far less attention.

Part of the challenge is that data governance is neither “fun” nor “sexy.” It lacks the excitement of new technologies and the appeal of quick wins, so it is consistently deprioritized.

Yet as organizations scale their AI ambitions, governance is increasingly the factor that determines whether those efforts succeed or stall.

The imbalance in attention is now starting to show. While AI adoption continues to grow, many organizations still struggle to move beyond pilot stages into enterprise-scale deployment. The gap between ambition and execution is widening, and weak data governance is often at the center of it.

The issue is not awareness. Most business leaders recognize that governance matters. The challenge is that governance demands structural decisions, cultural alignment and sustained discipline – the hard parts of the job. And, unlike a new platform or tool, its value often only becomes fully apparent when it is missing.

When governance is absent, problems don’t stay small

Weak governance rarely fails loudly at first. The problems accumulate.

Early AI initiatives often prioritize delivery, with dashboards, models and applications taking precedence over governance. Silos form, data definitions diverge and access controls become inconsistent. A common pattern: two teams – one in marketing, one in data science – train separate models against different definitions of the same metric.

Both definitions look correct in isolation. In production, the predictions conflict, neither team can explain why, and the investigation takes weeks longer than building either model did. Quality issues are patched rather than fixed, and new projects begin to rely on shaky assumptions.

As complexity grows over time, confidence in the data declines.

Data dictionaries and permission frameworks are not administrative overhead – they are what makes scalable AI possible. Building them early demands investment before visible returns but postponing that effort is far costlier.

Left unchecked, governance gaps eventually land hard, resulting in delayed projects, compliance failures and decisions made on unreliable data. At that point, organizations are forced into reactive fixes – or even total rebuilds – that are far more expensive and disruptive than addressing governance from the start.

Governance is not just compliance – it enables innovation

Regulators are placing increasing importance on accountability in how data is used. The UK’s Information Commissioner’s Office (ICO) has made it clear that organizations must be able to demonstrate control over data use, particularly as AI systems become more prevalent. Scotland’s new National AI Strategy also highlights that organizations must follow best practice in responsible AI governance aligned with OECD principles.

This has reinforced the perception that governance is primarily a compliance exercise – something important but not necessarily prioritized at the prototype stage. Effective governance is far more than that: it shapes how data flows through an organization, how decisions are made and how confidently teams can act. It defines accountability and sets the standards needed to maintain consistency at scale.

In that sense, governance is a design choice, and businesses need to make the right one to effectively scale their innovation ambitions.

Define ownership before you decide the model

Governance is not one-size-fits-all - nor it is purely a technical problem to be addressed through tools or platforms alone. In fact, the harder initial challenge is often a people and accountability one. Before designing a governance model, organizations need to define the who as much as the how. Who owns the data? Who is responsible for its quality and who decides how it should be used?

In many organizations, these responsibilities are unclear. Management is shared, and ownership is (often wrongly) assumed rather than defined. But it is only once those questions have been answered – in practice as well as on paper – that businesses can turn their attention to developing a governance model that fits their structure.

Some take a centralized approach to this, with control sitting in a single function. This can provide consistency and clarity, but the model may struggle to scale across complex organizations with diverse needs.

Others adopt a federated model, combining central standards with local ownership. This can be more flexible and scalable, but only if the business is committed to those shared standards and has defined clear roles and accountability. Without them, federated models risk furthering data fragmentation.

The key is alignment. Governance models should match how teams actually use data and AI, not how they’re assumed to operate.

A practical test: ask three different teams how they define a key business metric – revenue, active users, or customer churn. If the answers differ, the governance problem already exists. The operating model question is not how to prevent that divergence in future; it is who has the authority to resolve it now.

Governance doesn’t show up in a demo

Governance is rarely the most visible part of an AI strategy. It’s detailed, structural work that often goes overlooked, but that is precisely why it matters.

For business leaders, the challenge is to move beyond acknowledging its importance and begin making early, deliberate decisions about how it is implemented. That means defining data ownership, aligning operating models and investing in the capabilities that support long-term success.

Technology choices are reversible. Data ownership decisions compound. The governance model you design – or neglect – in the next twelve months will shape what your AI strategy can actually deliver in three years.

We've featured the best small business software.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Getting AI-ready while building your data infrastructure is like learning to drive a manual transmission on the wrong side of the road.

It’s complicated and requires potentially dangerous multitasking.

Organizations with immature data-handling processes that are adopting AI are trying to solve multiple technology problems at once, and risk stalling out.

Unsurprisingly, 48% of enterprises cited data-related issues as their top challenge to AI adoption in NVIDIA's 2026 State of AI report.

Most enterprise AI programs don't fail because of the model or solution selected. They fail because underlying data is fragmented, inconsistent and poorly governed.

Get Your Data Foundation in Order

Enterprise data is messy in layers. It’s scattered across many systems, making it hard to pull together into a coherent picture. Even when you can consolidate it, you often will run into granularity or identifier mismatches. One application may store account numbers as plain digits, while another adds “ACCT” as a prefix. That small inconsistency creates an extra reconciliation step every time you join those data sets.

Data governance compounds the problem. Without a system intentionally designed to control who accesses data, where it moves and what protections are in place, gaps emerge fast. PII exposure is the most obvious risk: an email address that ends up in the wrong hands can trigger a serious breach. Raw, unstructured data also yields mediocre AI outputs and is more expensive to process.

Clean, structured data yields better results at lower cost. A third gap, explainability, is quickly becoming a legal requirement. Many countries and several U.S. states now require organizations to demonstrate how AI-driven decisions were reached. Cut corners on the data foundation and you may not be able to show that chain of reasoning.

At that point, you’re either in compliance violation territory or your model is producing outputs you can’t defend.

Three Steps to Get Your Data AI-Ready

Define governance before you deploy. Classify your data: what is it, where did it come from and who can touch it. Separate the roles of technical decision-making and compliance oversight. Keeping those responsibilities with different people prevents a compromising situation where the same person sets the rules and monitors compliance.

Run cross-functional AI governance as a standing function. Assign a representative from every department and meet monthly to discuss what teams are working on, what concerns have surfaced and what support they need from one another.

Approach larger AI-readiness initiatives like any other business project: assign a project manager, designate an executive owner, set a weekly cadence, build a task list and work through it.

Collect behavioral data even before you need it. The outcomes you get from AI vary enormously depending on how skilled the operator is, ranging from using it as an expensive search engine to developing autonomous workflows. Without visibility, you might be pouring money into AI licenses and getting Google-level output in return.

You don’t know who needs training, whether they have the right tool in front of them or what outcomes they’re achieving. The risk is that you make the wrong strategic call as a result—abandoning a rollout, for example, when the real fix was better training or a different tool.

Further considerations

Here’s another layer to consider. When an experienced worker completes a task, with AI assistance, they leave more skilled than when they started. The output and the learning happen together. That's what behavioral data should demonstrate over time – not just task completion, but upward skill trajectories.

When someone at the beginning of the learning curve accepts whatever AI produces without critically engaging with it, you get the output but not the growth. Behavioral data is how you catch that gap early, before it becomes a long-term cost you can't unwind.

Stay curious and look for the easy wins. Focus your data readiness efforts on the workflows where work actually happens, and prioritize tools that let you get at that data.

A recent example illustrates the payoff. A product manager ran an AI-powered analysis of quarterly bug patterns using data from the department’s most commonly used tools. The results were unexpected. One team carried a disproportionate share of incoming tickets, most of them requests for manual workarounds to a missing product feature.

While other teams split their time roughly 75% on new work and 25% on incoming bugs, that team was closer to 50-50. By not building a single feature, the organization was effectively operating 1.5 people below capacity.

The entire analysis took about 45 minutes. None of it would have been possible without data that was organized, tagged by team, connected to individual contributors, accessible via existing AI connectors and protected by role-based access controls.

The organizations that get the most from AI are the ones that empower their people to ask "I wonder if there's something here" — and have data to diagnose in an afternoon. That only happens when the foundation is already in place.

Make your data safe with the best backup software.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Recent court developments in Germany have put an important issue into the robotics spotlight.

A German court in Hamburg has issued a preliminary injunction against Elite Robots Germany in a copyright infringement case involving copying of Universal Robots software.

As a result, the company is not allowed to offer or distribute the products covered by the decision in Germany while the case continues.

At first, this may sound like technical legal news only relevant for the German market.

But it highlights broader questions that matter to every company investing in automation – especially when choosing a collaborative robot that will operate close to people and become part of daily production.

Here are five reasons why.

1. Copying creates real risk for customers

When protected robot software or design is copied without permission, the impact extends well beyond the supplier and exposes all parties in the value chain to significant legal risk.

It can affect end-customers directly as using an infringing product for commercial purposes, such as a robot with infringing software in a production line, can itself constitute a legal violation.

This not only creates a risk of court-ordered remedies, including preliminary or permanent injunctions requiring the immediate shutdown and removal of the affected robots, but also exposes customers to costly and disruptive litigation and potential business interruption.

Automation systems are long‑term investments meant to run for years. Legal uncertainty at supplier level can turn into a real business risk on the factory floor.

2. “Similar” does not mean safe

Collaborative industrial robots are often described as safe, but safety is not automatic. It depends on how a robot is designed, tested, and used in real applications.

A robot that looks or behaves like another system does not share its safety profile. Safety comes from reliable hardware, validated software, certified functions, clear limits, and proper documentation. These cannot be copied by appearance alone.

Superficial similarity creates a dangerous false sense of security, which may result in serious physical injury to operators and bystanders.

3. Lower price can mean higher cost later

The purchase price of a robot is easy to compare. The long‑term cost is not.

If your robot vendor ends up in a legal battle, besides the question of even being able to use it legally, you also face uncertainty about product availability, software updates and service support.

Unexpected downtime, lack of updates or compliance challenges can quickly outweigh any initial savings. This has never been more relevant as modern robots are software‑driven machines. Motion control, force limits, diagnostics, and safety logic all depend on software.

If customers do not know where the software comes from, who owns it, or how it is maintained, they introduce uncertainty into production. Original, well‑understood software is essential for reliable and predictable operation over time.

In automation, shortcuts often appear affordable at first but expensive later. And as with all things in life: if it seems too good to be true, it probably is.

4. Buying copycat tech shapes the future of automation

Every automation investment sends a signal about what the market rewards. Choosing original, lawfully developed technology encourages long‑term engineering, robust safety practices, and continued product improvement. Choosing copycat technology does the opposite: it normalizes shortcuts, weakens incentives to invest in research and compliance, and shifts competition away from quality and reliability.

Over time, widespread tolerance of intellectual property infringement affects the entire robotics ecosystem, from suppliers and integrators to suppliers and regulators. It increases uncertainty and ultimately makes it harder for manufacturers to rely on stable platforms that will be supported and improved for years to come.

Protecting original technology is not about limiting choice or slowing competition. It is about ensuring that competition is based on real innovation, verified safety, and accountability – and that customers can invest in automation with confidence, knowing the technology they rely on is built to last.

5. Trust in the original

In summary, choosing a robot is not only about specifications and price. It is about trust.

Buyers need confidence that a robot is legally sound, properly certified, and supported by people who truly understand the technology. Trust comes from transparency, responsibility, and deep technical knowledge – not from claims or visual similarity.

Automation is becoming increasingly central to modern manufacturing and as it does, questions of originality, safety, and integrity become part of responsible decision‑making.

So, no matter if you’re purchasing your first robots or expanding your fleet, before asking what a robot can do, it is worth asking a simple question:

Do you trust where it comes from?

I tried 70+ best AI tools.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

One of the most significant changes in the last couple of decades has been the rise of social media and the way it's reshaped society. Mark Zuckerberg, who co-founded and launched Facebook – and now leads the technology conglomerate Meta – has played a key role in what he describes as an erosion of privacy.

Social media shifts

We were just months away from Mark Zuckerberg's first on-screen portrayal in The Social Network when he spoke about the story of Facebook so far at TechCrunch's Crunchies awards in January 2010.

During the interview, he spoke about the way that people's attitudes toward their own sense of privacy had shifted, and that with the tools available on the web – including his own platform – users were far more open and willing to share details about themselves.

He delivered these thoughts following criticism that Facebook had altered its architecture so that basic information, like your friends list, became publicly viewable by default – unless you consciously opted out.

The death of privacy

Zuckerberg, at the time, was attempting to reframe his platform's changes as ones that served users as it adapted to their behavior. But it's also fair to suggest that Facebook had also played a role in triggering behavioral shifts in its users in the first place.

Cynics would also suggest that the shift to opening up more data to public access paved the way for further monetization of user data. This would come in the form of Open Graph, which the company launched in May that same year.

People have continued to share their information online, but they aren't feeding simple algorithms or monetization portals, but AI models too.

On the other hand, the modern age has given rise to more encrypted services like WhatsApp, Signal and Discord, with mostly open platforms, like Facebook's previous incarnation, suffering from data collection scandals and privacy breaches.

I’m under no illusions that John Ternus’ Apple will be the exactly same as it has been under Tim Cook. Initially though, there will be few, if any, changes. Ternus is not an outsider with wild anti-Apple ideas intended to wake a sleeping giant. He's been here for decades, through all the major releases that made Apple, well, Apple. Claims that he's arriving in September to revive Apple's design excellence are, if not off base, then just wrong-headed.

First of all, the argument presupposes that there is something fundamentally lacking in Apple's Industrial design: Jony Ive was obviously lighting in a bottle, and current design lead Molly Anderson is a pale, albeit also British, imitation. (Granted, Anderson has not been in the position that long, taking over from Evans Hankey, who left in 2023.)

But accepting that presumption is to ignore all the beautifully designed products that have arrived under Cook's leadership, with and without Ive, and also often under the watchful eyes of Ternus.

Remember the Pro — oh

But since Ive's skills as a designer are so vaunted, let's start with a failure. Perhaps you remember the Mac Pro. No, not the cheese-grater design unveiled in 2023. While that had its detractors, it was miles above the trashcan design unveiled in 2013.

It was the ultimate expression (at least in PC terms) of Ive's "form meets function" obsession. The internal structure was sort of a triangle of boards that seemed perfectly wrong for a squat, circular enclosure. Ive's fingerprints were all over the impractical system, one that Craig Federighi later admitted to me (during a mea culpa meeting on the Pro) had boxed Apple into a thermal corner. It was hard to upgrade and was roundly rejected by pro system users.

Ive was also responsible for the Apple Pencil. With it, his penchant for skeuomorphism extended from app design into the physical world. The Apple Pencil fully resembled a white plastic version of a real, pencil, and to accommodate that, it had, under a custom cover, a hidden Lightning charging plug. You even needed a special adaptor to charge it. Later, wirelessly charging Apple Pencils, which I think Ive also designed, fixed this mess.

Naturally, Ive's hits far outweighed his misses, and many big ones came during the Cook-Ive collaboration period, including the iPad Air, the Apple Watch, and AirPods (which started awkwardly but gradually improved).

The more divisive Apple Vision Pro was likely designed under Hankey and Anderson's watchful eye. I do think it's a pleasant intersection between the needs of extremely high-end innovation and aesthetic appeal. Goggles will be goggles, after all.

Even iPhone design has remained, if not excellent, interesting.

The iPhone is still beautiful...discuss

When Apple introduced the first iPhone in 2007, there really was no other handset quite like it. Apple set the bar and, over the years, every other manufacturer followed. It became harder and harder for Apple to differentiate its metal and glass slab from its competitors. The ever-larger camera arrays have provided a sort of design challenge and an opportunity at the same time. Hankey and Anderson, at least, came up with a giant island that's quite recognizable from a distance. And let's not discount color. The orange was a stroke of genius and the new MacBook Neo's playful Blush and Citrus colors are lively, proving Anderson knows how to marry form with expression.

Ternus' role as hardware lead means he's been seeing these designs for years. My sense in sitting down with him after the launch of the iPhone Air is that he is intimately involved with the process of shoving all that technology into ever-thinner, but shockingly strong frames. He gets that you need to marry engineering skills with industrial design to get a durable and usable product that's still attractive.

When I think about what Ternus will do when he finally takes over in September, I am reminded of relay race runners. Cook continues to pace around the track while slowly holding out the baton behind him. Ternus is nearby, running just behind Cook with one hand outstretched. Neither man will stop. The handoff will happen in a few months, with everything still in motion.

The process of finishing the rumored iPhone Ultra (the folding phone) continues as we speak, and Ternus is not going to suddenly pull a Steve Jobs and throw the design into a fish tank, to see if bubbles rise, thus proving they could make it smaller. He won't demand that all plastic be replaced with glass, or even that one more color be added to the mix.

When, as the recent Bloomberg report claims, Ternus said of Apple's design history and appeal to customers, "We’re going to make sure that stays the case,” he's not talking about making huge changes to achieve that goal.

Ternus will stay the course and support Anderson. He may hire more design support and, down the line, Ternus will look for his signature initiative or product; he will want to have his own iPhone or iPod. But that's a natural inclination for any incoming CEO. Job one, though, is staying the course, shepherding the in-the-pipeline products to market and ensuring that Apple remains Apple.

I think Ternus knows exactly how to do that.

Professionals end every day feeling behind or burnt out, but not because they haven’t worked hard enough or clocked off earlier that day.

It’s because the volume of information, decisions, and context-switching is moving faster than the pace humans can realistically handle.

A recent Microsoft report put numbers to what people are feeling.

Eight in 10 of the global workforce say they lack enough time or energy to do their work, and 60% of meetings are happening as ad hoc calls or quick chats outside the pre-scheduled day-to-day.

This isn’t a motivation problem, it’s a capacity one - and it's created one of the defining contradictions of modern work. Businesses have never had more ideas, expertise, or ambition at their disposal, yet the people inside them are increasingly starved of the time and clarity needed to turn that potential into progress.

The smartphone makes this contradiction impossible to ignore. It is one of the most consequential inventions of the 21st century, yet also one that many people actively try to use less. Screen-time limits and digital detoxes are not anti-technology trends. They are signs that people are trying to regain control over a tool that has become indispensable, but increasingly overwhelming.

The message is simple: the market isn't asking for more technology. It's asking for relief.

Technological exhaustion

People are adopting or looking at things like digital assistants, wearable AI, focus apps, and workflow automation, not because the technology is impressive. They're doing it because they're exhausted.

That distinction matters because cognitive overload has become a workplace crisis. And the first wave of wearable AI missed the opportunity to solve it.

Instead of building practical tools, companies chased futuristic visions. Early wearable AI products asked "what can AI do?" instead of "what problem needs solving?"

The Humane Pin is the most obvious and probably the most well-known industry example. The vision was compelling, but the execution wasn't there. It positioned itself as a complete phone replacement before proving it could do even just one thing better than a phone. Ultimately, it tried to be everything and ended up being nothing.

This approach didn't reduce cognitive overload - it created more. Another device to manage. Another thing running in the background of an already overwhelming life.

The wrong question asked was: "How do we replace the phone entirely?"

A much better question is: “Where are people losing the most time, energy, and clarity — and how can technology give some of it back without demanding more from them?”

Useful technologies

The most useful technologies rarely arrive by replacing everything at once. The calculator didn't try to replace the accountant - it eliminated one specific source of friction and became indispensable.

It’s the same with wearable AI assistants. Progress is made in practice, not promises.

The wearables gaining real traction share one quality: users can explain their value in a single sentence. "This device exists so I can stop worrying about X." That clarity isn't a constraint - it is the product.

The future of this category will not be defined by the devices with the boldest premise. It will be defined by those who understand where people are most overloaded and remove that pressure without asking for much in return.

Does the technology make someone feel more capable or more managed? Does it reduce the number of things they have to remember, check, repeat, and translate? Does it create clarity, or simply another stream of information?

Those questions are less glamorous than asking whether AI can replace the smartphone. But they are also far more useful.

The wearables that will actually help aren't the ones with the boldest premise; they're the ones that solve one real problem but do it well.

In a world drowning in information, that may be the most ambitious thing technology can do.

Simplify work with the best AI tools.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

For most of the world, the 2026 FIFA World Cup will be remembered as a sporting event. For cybersecurity teams, it will function more like a live multinational stress test.

Spanning three countries, 16 host cities, and thousands of miles of transportation corridors, the tournament depends on an interconnected ecosystem of physical and digital infrastructure operating under sustained pressure for more than a month.

Airports, rail systems, hotels, fan festivals, credentialing platforms, broadcast operations, rideshare services, and public-facing digital services will all be strained simultaneously.

That scale fundamentally changes the security equation.

From a threat intelligence perspective, the defining challenge of the 2026 World Cup is the convergence of physical, cyber, social, and geopolitical risks across shared IT infrastructure and compressed operational timelines. Security teams are no longer managing isolated threats — they are forced to manage cascading disruption, where pressure in one domain can rapidly affect another.

A phishing campaign targeting transportation staff could disrupt rail operations moving tens of thousands of fans. A localized protest could overwhelm nearby transit systems and alter executive movement plans. A ransomware incident affecting a hospitality provider could create physical security concerns if communications or access systems fail during peak crowd periods.

This is the reality of large-scale global events in 2026: the attack surface is no longer just the venue, it’s the infrastructure surrounding the whole event.

At the time of writing, Flashpoint has not identified any specific, credible threats targeting the tournament. That should not be mistaken for a low-risk environment. Events of this scale consistently attract opportunistic criminal activity, fraud operations, extremist messaging, coordinated protest movements, and attempts to exploit operational strain.

The Security Perimeter Extends Far Beyond the Stadium

Historically, security planning for major sporting events has centered on venue protection. For the 2026 FIFA World Cup, that model is no longer sufficient.

With matches spread across three countries and 16 cities, much of the risk now sits outside controlled environments across transit systems, hotels, fan zones, entertainment districts, and the broader infrastructure moving people, information, and services between them.

In many cases, these environments carry greater uncertainty than the stadiums themselves. Security visibility is uneven, access controls are inconsistent, and crowd density, alcohol consumption, and movement constraints create conditions where relatively minor incidents can escalate quickly.

Protests are likely to add another layer of complexity. Demonstrations tied to immigration policy, labor concerns, geopolitical tensions, and broader political movements are expected to occur across multiple host cities during the tournament. Most demonstrations will likely remain lawful and localized.

While most demonstrations will likely remain lawful and localized, the risk emerges when protest activity intersects with transportation choke points, fan movement patterns, or already strained public infrastructure.

Threat intelligence teams should pay close attention to how online rhetoric translates into physical coordination.

Many of the indicators that matter most during events like the World Cup appear early through fragmented digital activity: encrypted messaging channels, localized social media conversations, extremist propaganda ecosystems, fraud marketplaces, and open-source coordination efforts.

The intelligence challenge is rarely a lack of data. It is identifying which signals indicate a meaningful shift in operational risk.

Crowd Dynamics and Operational Disruption

Crowd behavior remains one of the most persistent, and often underestimated, security challenges at large-scale events.

Mass gatherings create conditions where panic can spread faster than verified information. Overcrowding, pyrotechnics, aggressive supporter behavior, or sudden movement within confined transit areas can trigger cascading safety incidents without any organized attack occurring.

Recent years have also shown increasing coordination among certain supporter networks and hooligan groups, including the use of encrypted communications and reconnaissance activity to organize around less-secured gathering points outside official venues.

These risks matter because they place pressure on the systems surrounding the event, not solely the event itself. The same convergence is visible across the cyber threat landscape.

We are likely to see elevated levels of phishing activity, ticket fraud, domain impersonation, social engineering, and opportunistic attacks targeting tournament-related infrastructure. Threat actors understand that large events create urgency, emotional decision-making, and predictable behavior patterns. Fans searching for tickets, transportation, accommodations, or livestreams become easier targets for spoofed domains and fraudulent communications.

The operational implications extend well beyond consumer fraud losses.

A disruptive cyber incident affecting transportation systems, hospitality providers, third-party vendors, or venue operations during a high-attendance match day can rapidly create downstream physical security challenges. Delayed transit systems increase crowd concentration.

Failed communications systems complicate emergency response coordination. Access-control outages create confusion at security checkpoints. Small technical failures can compound quickly in dense environments operating on fixed timelines.

What Security Teams Should Prioritize Before the Tournament

Organizations supporting personnel, executives, vendors, or operations during the World Cup should prepare for an environment where physical and digital disruptions increasingly overlap.

That preparation starts with visibility.

Security teams should establish continuous monitoring around transportation disruptions, protest coordination, fraud infrastructure, and emerging operational incidents across both open and closed online sources. Threat indicators tied to major events often surface first through fragmented local reporting, encrypted messaging channels, social media coordination, and opportunistic criminal communities.

Travel security planning should also extend beyond venue access and hotel bookings.

Organizations should review how employees handle credentials, travel itineraries, executive movement, and event-related content online. During high-profile international events, threat actors routinely collect publicly available information to support impersonation attempts, social engineering campaigns, and physical targeting.

Employees, contractors, media personnel, and attendees frequently expose operationally sensitive information online without recognizing the downstream implications. Credential badges, transportation routes, executive locations, hotel details, and backstage access procedures often appear publicly across social media within minutes. Threat actors increasingly use these fragmented disclosures to map security procedures, identify soft targets, or facilitate social engineering operations.

Third-party dependencies deserve particular attention.

Hospitality providers, transportation vendors, temporary staffing organizations, event technology platforms, and local service providers will all operate under elevated pressure during the tournament. Security incidents affecting those organizations can rapidly create downstream operational disruption for attendees, sponsors, media teams, and corporate travelers.

Finally, security leaders should prepare for disruption scenarios that fall below the threshold of a major crisis but still create operational consequences. Delayed transportation, localized unrest, communications outages, credentialing issues, and short-duration cyber incidents can all affect executive movement, employee safety, and business continuity during compressed event timelines.

The Organizations That Adapt Fastest Will Be Best Positioned

Events like the 2026 FIFA World Cup place unusual strain on security teams because disruption rarely stays contained within a single domain.

A cyber incident can create immediate physical consequences. Protest activity can disrupt transportation and executive movement. Crowd-management failures can generate downstream operational strain across hospitality, communications, and emergency response systems.

For security leaders, the challenge is maintaining visibility across these interconnected environments as conditions evolve in real time.

The organizations best positioned during the tournament will not necessarily be those with the largest physical security footprint. They will be the organizations capable of continuously correlating cyber indicators, physical activity, online narratives, and emerging operational disruptions into a coherent picture of risk.

Threat intelligence creates decision advantage in environments where conditions evolve by the minute.

We feature the best internet security suites: ranked and rated by experts.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

• The Blair Witch Project reboot will be released on September 24, 2027
• YouTube creator Dylan Clark will direct the project
• Clark has already directed a number of horror shorts which are available for free on YouTube

The Blair Witch Project is getting a reboot, which will arrive in theaters worldwide on September 24, 2027. But of all the horror titles to reboot, I'm concerned this was a poor choice.

YouTube creator Dylan Clark has been announced as the director of the rebooted found footage horror movie, and there's no denying he's talented. His YouTube channel features a number of effective short films you can watch for free.

News of The Blair Witch Project's reboot was confirmed by Lionsgate on social media, where they shared the release date announcement.

Is a Blair Witch Project reboot really necessary?

That's the big question among horror fans right now, and I'm sure people have plenty of strong opinions about it.

I'm certainly excited for Dylan Clark to be working with Lionsgate, but in a world where YouTubers Kane Parsons' Backrooms and Curry Barker's Obsession are dominating the box office with their original stories, I'm unsure we should rely on a YouTuber to reboot an iconic movie like The Blair Witch Project instead of creating something new.

Indeed, recent box office figures have shown that there's an appetite for original storytelling, with both Parsons and Barker receiving widespread acclaim for their horror movies.

Backrooms made $277.4 million from its $10 million budget, and Obsession made $334.4 million from a budget of just $750,000 - incredibly impressive results which have no doubt caught the attention of Hollywood.

It's clear that horror fans want something new, and I'm among them. The only way I can see The Blair Witch Project working is if it is completely stripped back and told from a new angle.

But even so, perhaps a completely new found footage horror would have turned out better? When the original movie came out in 1999, it was exciting because it was an entirely fresh concept.

Curry Barker is currently working on a Texas Chainsaw Massacre reboot, which I was initially apprehensive about too, but he confirmed that he would be focusing on lesser-explored characters. I truly hope that Dylan Clark takes a similar angle and offers something new when it comes to The Blair Witch Project.

Right now I'm not filled with much confidence, but as always I am open to being proved wrong.

Enterprises built their early AI strategies in a world that assumed relative freedom of data movement. That world no longer exists, so how do they adjust to the new world?

While it’s a cliché that data is the new oil, the truth is that data does need to be able to flow around an organisation and come to rest in some central location, where it can be refined to extract insight and value.

In recent years, that refining has come, of course, via AI.

But if we really think that analogy through, it also captures some of the problems enterprises currently face managing data.

Recently, the world has seen what happens when the free flow of oil, gas, and other fundamental resources is disrupted. Governments and companies are scrambling to work around supply chain interruptions in the short term and looking for energy sovereignty and resilience in the long term. In this case, centralization has become a liability, thanks to a single, but critical, chokepoint.

In the same way, the upheaval created by AI, paired with governments’ and regulators’ attempts to manage its impact, means companies need to rethink how they architect and manage their data. The difference is that the ways traditional data flows around organizations are changing.

After all, a centralized approach made sense when moving data was straightforward. Now, however, governance and sovereignty concerns and data movement costs have changed the calculus.

AI transformation

AI alone has transformed the equation when it comes to the volume of data that must be managed. It’s not just training models that require massive amounts of data. Those models need constant updating and tuning with fresh data. And, if those models are to deliver value to the business, companies will be constantly running inference, which requires more data, and generates more data.

Unsurprisingly, governments and regulators have an interest in AI governance. This includes existing concerns about data residency and the possibility of new problems, such as data leakage into LLMs.

In Europe, for example, the regulatory landscape laid out by the GDPR mandate has been complicated by the rollout of the EU AI Act, which comes fully into force from August this year. In the US, companies face multiple federal and state-level regulations. Any, or all, of these could come into play as data is moved to that central refinery.

Governments also have an understandable strategic interest in developing sovereign AI, further complicating matters when it comes to both data and the models that work on it.

Once we consider this, it’s clear that only a few organizations are architected to handle the sheer scale of data involved in AI, and the governance it requires.

Companies grow messily, whether organically or through M&A, inheriting different infrastructures – and different governance regimes - as they sprawl across borders.

Data in the old world

In the old world, free-flowing data might have been seen as efficient. Now, given the vast amount of data involved, generating copies of data and moving data across borders is both fraught with risk from a governance perspective. And with escalating cloud egress fees, data movement becomes extremely expensive. At the same time, moving or copying on-prem presents both a financial challenge, and imposes a burden on already stretched technology teams.

What are our options in this new world? Few technology leaders have the option of simply opting out of the whole AI revolution. Whatever your personal views on the technology, few C-suites are prepared to sidestep the AI race.

But we can lay out a roadmap for how to manage data – and compute – in this new world.

To start with, we need to understand the environment we’re really operating in. This may well include accepting that a hybrid or multi-cloud architecture is going to be the normal state of affairs. The very nature of AI, with companies needing to access multiple models and multiple services, means that traditional monolithic, central approaches simply won’t scale.

This, in turn, means every technology leader needs to be crystal clear on what governance means for their organisation and its data, whether that’s data privacy or residency requirements. And it’s imperative that governance is embedded in the AI workflow from the outset. It’s too important to be an afterthought or kicked down the road as other priorities arise.

What this makes clear is that it will normally make more sense to bring compute and those critical AI models to the data, not the other way round.

This isn’t just about reducing data movement costs or producing multiple copies of data. It’s about reducing the friction that comes with reconciling governance requirements as data moves around organizations and across borders. But it can also mean reduced latency and fresher data, making AI more effective.

That’s not to say that it’s not sometimes necessary to move data. But if that’s the case, let’s be mindful and deliberate about it.

But while data becomes increasingly decentralized, it’s imperative that we centralize the management of data access and build the platform accordingly. This lays the groundwork for clear data governance and sovereign AI alike.

By rethinking how they manage data movement, technology leaders can bypass the escalating egress costs and compliance traps embedded in the old way of doing things. The result is an AI strategy that is both scalable and sustainable, enabling enterprise-grade AI, wherever their data lives.

We feature the best data migration tools.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Most executives have no idea how much of their website traffic comes from AI agents.

If you were to ask which AI agents are legitimate and which are impersonating trusted names to scrape data, they’d struggle to tell them apart, a problem that’s growing by the day.

In early 2026, AI and bots generated billions of requests, outpacing internet traffic from humans.

This is no longer a fringe activity; AI agents are now a persistent, substantial portion of the traffic hitting websites.

Yet most organisations can’t tell you what that traffic is doing, where it’s really coming from, and whether it’s helping or hurting their business.

The Volume Trap

When organisations hear that AI agent traffic is creating billions of requests, the instinct is often to treat it as a monolithic category. It’s not. Lumping all AI agents together is like treating all humans as identical users; it misses the nuance that determines value.

Take two agents from the same company: one built to improve search relevance, potentially driving referral traffic back to a website, and another designed purely for large-scale data extraction to train AI models, offering zero benefit to organisations.

Both show up in traffic reports, both generate similar volumes, but only one has any upside for businesses. Without the ability to distinguish between them, companies can’t make informed decisions about either. Organisations are flying blind, and the cost of that blindness is steep.

The Trust Problem

Here is where it gets trickier: even when an AI agent identifies itself, organisations can’t trust it. Recent data shows that well-known, trusted AI agent names are being actively impersonated at scale. Meta-ExternalAgent was spoofed over 16 million times in early 2026. ChatGPT-User saw nearly 8 million fraudulent requests using its name. PerplexityBot had nearly 2.4% of all requests claiming to be legitimate turn out to be fake.

If website allowlists – approved lists granted automatic access - certain AI agents by name, assuming they are legitimate crawlers, a fake agent string is essentially a skeleton key. Bad actors know this and are using trusted agent identities as cover to bypass defenses and extract whatever data they want.

The exposure isn’t theoretical. Testing across 700k high-traffic websites revealed that the vast majority return full access to spoofed AI agent requests with no verification whatsoever.

The Agentic Browser Challenge

Traditional AI crawlers are only part of the story. A newer, more sophisticated vector is emerging: agentic browsers. These tools don’t just request a page, they simulate full browser sessions and interact with a site like a human user.

They’re harder to detect and harder to distinguish from legitimate traffic, and they are showing up in force across the industries with the most valuable transactional data.

In February 2026, agentic browser traffic was concentrated in ecommerce and retail (about 20% of volume) and travel and tourism (15%). These sectors hold some of the most valuable transactional data on the internet: pricing data, inventory information, customer behavior patterns, and competitive intelligence.

For businesses in any of these sectors it’s time to start actively monitoring for agentic browser activity, as organisations may be leaking data without realising it.

What This Means for Decision Makers

The implications of this visibility gap are immediate and material. Invisible traffic is unmanaged traffic. Companies that can’t identify traffic can’t decide what to do with it. Should it block it? Throttle it? Allowlist it? Monetise it? Without clear visibility, decisions become guesswork.

High volume does not equal high value. Some AI agents drive search visibility and referral traffic. Others extract data and contribute nothing in return. By treating them the same, organisations are subsidising data collection efforts with no upside for a business.

Relying on basic bot detection doesn’t cut it anymore. Agentic browsers behave like real users and simple signal-based detection misses them. Organisations need behavioural analysis that accounts for session patterns, timing, interaction signatures, and other contextual indicators.

Where to Start

Getting control of AI agent traffic starts with visibility. Organisations need to log and classify what is hitting sites, by agent type, behaviour, and claimed identity without relying solely on user-agent strings, as they’re easy to spoof.

Agent classification is an ongoing practice. As the AI agent ecosystem evolves quickly, with new agents appearing regularly and existing ones changing behaviour, in-time assessments go stale fast.

Establish a tiered access framework, but make it session-specific, not agent-specific. The same AI agent can exhibit legitimate behaviour in one session and extractive behaviour in another.

Intent-based detection evaluates what an agent is doing in real time, not just what it claims to be. Is it browsing product pages at a human pace or scraping an entire catalogue? The behaviour in each session should determine the response.

Companies should stop assuming that because something identifies itself as a known agent, it is legitimate. The cost of blind trust is too high. Verify everything.

AI agents are not going away. Their traffic will continue to grow, and their behaviour will continue to evolve. The organisations that thrive in this environment will be the ones that can see clearly what is happening on their websites and make deliberate, informed decisions about what to allow and what to block.

Right now, most organisations can’t, and that needs to change. AI agents are already interacting with websites. The question is whether organisations know what they’re doing while they’re there.

We feature the best website builders.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Tim Cook has long taken a strong stance against the infringement of Apple users' privacy – and the general erosion of privacy. That's been the case whether he's shown support for end-to-end encryption or if he's railed against the monetization of user data.

The rise of data protection

Almost a decade ago, the European Union (EU) introduced the most radical reformations to data protection laws in a generation with the General Data Protection Regulation (GDPR).

Several months later, the (now outgoing) Apple CEO spoke at the 40th International Conference of Data Protection and Privacy Commissioners with a speech that targeted Apple's fellow technology rivals with both barrels.

He pointed out that billions of dollars were changing hands – and countless decisions were being made based on data points harvested from our interactions on digital platforms. These may include clicks of a Like button but also the information we have shared, often without understanding the full implications.

Your very own digital profile

Cook projected a dystopian future in which each person would be represented by a digital profile that's been devised based on analysis of the countless data points systems have gathered.

The purpose of this form of behavioral profiling, he suggested in his speech, could range from more effectively monetising your information to targeting you with more extremist content in one direction or another.

For example, we've since seen the way that social media platforms and similar sites have been highly effective in populist political movements, including the U.K. 'Brexit' decision to leave the EU. There are also fears this sort of power has been weaponized, to use Cook's phrase, by foreign adversaries.

Nearly 10 years on from the introduction of GDPR, there are fears that the rise of AI – which is turbocharging some of the fears the outgoing Apple boss raised – is undermining the laws and that newer, more modern regulations are needed.

I’ve tested a lot — and I mean a lot — of the best Bluetooth speakers. More than 50 to be exact. And if we’re talking about sheer popularity, then one brand is the clear and obvious top dog: JBL. For good reason too. In my years of testing, I’ve found that JBL regularly delivers the ideal blend of quality and affordability, regularly producing Bluetooth speakers with great sound and impressive durability at an easy-to-stomach price.

JBL also makes loads of different options, so I though I'd help you narrow things down a bit, and have picked out three JBL speakers I’d actually spend my own money on — with speakers for all budgets.

I’ve selected a small, budget-friendly speaker, a mid-priced maestro, and a premium pick that’s ideal for parties — there really is something here for everyone. If a speaker hasn't make this list, it’s by no means a vote of no confidence; this is simply a list of the three I’d personally grab right now.

1. JBL Go 5

Let’s start with the small yet mighty JBL Go 5, which I recently reviewed and rated five stars. The Go 5 is, in my view, the best small speaker on the market right now, offering clear, precise, and well-balanced sound, alongside an extremely hardy build, and a stellar feature-set.

As a mini speaker, the Go 5 won’t be able to belt out earth-shaking bass, but it plays to its strengths. Low-end sound is agile and punchy rather than ‘boomy’ and uncontrolled. Meanwhile, mids are clear and composed, and treble is highly articulate, resulting in great sound within a small package.

There’s also USB-C audio for lossless playback and EQ options to tailor audio to your specific tastes. The Go 5 has a more layered, full sound than its predecessor, the JBL Go 4, so if you own that model, I’d still recommend upgrading.

Another reason to opt for the Go 5 over the Go 4 is its design. The newer variant is equipped with edge lighting that makes listening to music even more immersive — especially in low-lit rooms if you want a bit of ambience. Waterproofing also got a boost to IP68, meaning the Go 5 is fully dustproof and capable of surviving a 1.5 meter dunking underwater for as long as 30 minutes — ideal for beach trips and pool parties.

Mix in solid battery life, Auracast compatibility, as well as a drop proof exterior, and the JBL Go 5 really is the full package — albeit a small-sized one. It’s typically available for $54.95 / £39.99 (about AU$75) as well, meaning it’s an absolute bargain.

2. JBL Flip 7

If we’re talking about value for money, then the JBL Flip 7 might just be the greatest Bluetooth speaker I’ve ever tested. Big statement, I know.

The reason for this is actually fairly simple: the sound it produces defies belief — not only based on the speaker’s low price, but also on its low footprint. Yes, the five-star Flip 7 is very compact, but it produces big, impactful sound, with strikingly impactful bass that demands your attention. But this doesn’t come at the expense of the rest of the frequency range. Mids are layered and intricate with vocals sounding especially well-defined, while treble sounds exert a level of expressiveness that’s beyond expectation.

Elsewhere, the Flip 7 carries over a lot of the Go 5’s greatest features, like an IP68 rating, USB-C audio, Auracast, and PlayTime Boost if you need a bit more battery life — though the standard 14 hours should get you a decent way. Sure, there’s no edge lighting on this one, but it's a clear step-up in terms of sound quality and power — it has a 35W maximum power output compared to the Go 5’s humble 4.8W.

At $149 / £129 / AU$179, the Flip 7 already overdelivers against its asking price, but I’ve seen it pop up on sale plenty of times, so keep your eye out for a sweet deal.

3. JBL Xtreme 5

Last but most definitely not least, I have the JBL Xtreme 5. If you’re working with a larger budget, want massive sound, or want a speaker for parties, then this one’s for you. This model was released alongside the Go 5, so I also reviewed it very recently, and like its small counterpart, it earned a glowing five-star review.

But what makes the Xtreme 5 so special? Of course it has all of the aforementioned features, whether that be IP68 dust and waterproofing, wired lossless playback or personalizable EQ. But the big difference is made in the audio department.

The Xtreme 5 has a massive maximum power output of 130W, meaning it can supply seismic sound, with thumping bass, driven and direct mids, and vibrant highs. Although it's a great performer all-round, it really is the low-end that wows. I described this model’s bass as “mesmerizing” in my review, and during testing I was blown away by the sheer might of its low-end output, which was significantly improved from the JBL Xtreme 4.

In addition, the Xtreme can squeeze out up to 28 hours of playtime, has JBL EasySing Mic compatibility for karaoke, and like the Go 5 it has customizable LED lighting. If you want the ultimate portable party speaker, look no further than this. The Xtreme 5 usually sells for $399.95 / £329.99 (about AU$560), which is very competitive in the upper echelons of the Bluetooth speaker market, although it's undoubtedly quite the investment.

For a long time, security teams have been dealing with the same problem: a constant stream of security alerts, but not enough context.

Missing details like user behavior, asset importance, or related activity, means there’s a heavy reliance on analysts to work out what actually matters.

This doesn’t just slow teams down; it puts real pressure on teams and limits how much they can realistically review or understand.

Agentic AI changes this dynamic.

Instead of looking at alerts in isolation, these systems can piece activity together, understand what’s happening in context, and in some cases take action on their own.

Often, issues are resolved before they ever need to be escalated. That removes a lot of the manual effort that’s shaped security operations for years.

But while a clear improvement, it doesn’t remove risk—it shifts it.

As systems improve, scrutiny declines

A useful comparison is aviation. As systems become more reliable, people naturally step back. Not because they’re careless, but because constantly double-checking something that’s almost always right starts to feel unnecessary. Over time, trust stops being something you actively think about and becomes something you assume.

The same thing is starting to happen in cybersecurity. As these systems prove themselves, teams spend less time questioning individual decisions. The environment feels calmer, and the lack of issues reinforces that sense of control. The real risk isn’t frequent failure, it’s that when something does go wrong, it’s less likely to be challenged.

Alert fatigue comes from having to pay attention to too much, too quickly. What follows is something different: a gradual drop in attention, where growing confidence in the system weakens the instinct to double-check.

A model built on two interdependent layers

The structure of security operations starts to shift as well. Instead of everything hinging on human decision-making, you end up with two connected layers. People set the intent – defining policy, access and boundaries – while agents interpret it and act on it, often much faster than any person could.

Both layers can be influenced. Traditional attacks aimed at people don’t go away, but there’s now another surface to think about: the data, prompts, and workflows that shape agent behavior. If those inputs are manipulated, the system can still produce actions that look valid, because they follow its internal logic.

At the same time, the distance between decision and execution increases. Human operators aren’t as involved in the moment an action happens, which makes it harder to spot when something isn’t quite right. In practice, each layer ends up relying on the other for validation.

When that assumption holds, the system works efficiently. When that works, everything runs smoothly. When it doesn’t, the gap between them can be hard to see in real time.

How risk scales in an agentic environment

Risk doesn’t just increase in this kind of environment, it spreads differently. Each agent has its own identity, permissions, and decision-making logic, and they’re often connected. Actions taken in one part of the system can trigger responses elsewhere, creating chains of automated behavior.

That means a single bad input or flawed decision doesn’t stay contained. It can move quickly across systems without anyone stepping in. The issue isn’t just speed, it’s how connected everything is. Small mistakes can have much bigger consequences because they’re carried through multiple layers of automation.

Why identity and access need to change

How agents are set up today adds another layer of risk. In many cases, they’re treated as extensions of the user, with the same credentials and access. It’s convenient, but it also widens the blast radius if something goes wrong.

A more resilient approach is to treat agents as their own entities. Give them distinct identities, limit what they can do to specific tasks, and make sure their actions can be tracked and reversed if needed, without affecting everything else.

It’s less about efficiency and more about putting the right foundations in place for systems that are increasingly acting on their own.

Maintaining control as reliance increases

One of the trickier aspects is that failure doesn’t always look like failure. Fewer alerts and faster resolutions can make it feel like risk has gone down, when in reality oversight may just be less active.

Staying in control comes down to how these systems are designed and used. High-impact actions still need some form of verification, even if most routine work doesn’t. It also matters that teams can see not just what an agent did, but how it arrived there—what inputs it used and how it interpreted them.

The ability to step in is just as important. If stopping or overriding an automated process is slow or awkward, it probably won’t happen in time when something goes wrong. That kind of intervention needs to be simple enough to use under pressure.

More broadly, the role of the security professional shifts. It’s not just about spotting obvious problems anymore, but recognizing when something that looks fine might still need a second look.

A quieter, more concentrated risk

Agentic AI will do a lot to reduce alert fatigue, which has weighed on security teams for years. The trade-off is that risk becomes less visible and more concentrated in the space between what people intend and what machines actually do.

In systems that work correctly most of the time, the real challenge isn’t constant failure. It’s what happens when something does go wrong and whether the usual signals that would catch it are still there.

We've reviewed, rated, and ranked the best firewall software.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

There is a growing tendency to frame advances in artificial intelligence through the lens of breakthroughs: new models, new architectures, new capabilities.

Innovation is often measured by what is invented, and how quickly.

But in some domains, this framing misses the point entirely.

Financial crime is no longer a collection of isolated incidents, it has evolved into an organized, adaptive, and increasingly industrialized system.

Criminal networks operate across geographies, leverage automation, and continuously refine their methods: they test, iterate, and scale, just as any high-performing organisation would.

In such an environment, the question is not whether an AI system is innovative. It is whether it can operate at the same level of organisation, speed, and adaptability as the threats it is designed to counter.

When Innovation Meets Reality

Most AI breakthroughs do not survive contact with real-world systems.

In controlled environments, models perform well. Data is structured, assumptions hold, and evaluation metrics are stable. But reality introduces a different set of constraints: incomplete information, shifting behaviors, latency requirements, regulatory oversight, and adversarial actors actively attempting to exploit weaknesses.

Financial and state systems, in particular those related to fraud and risk, represent one of the most demanding environments for AI. Decisions must be made in milliseconds, errors carry direct financial and reputational consequences, and the underlying patterns are constantly evolving. Not randomly, but intentionally.

Fraud is not a static problem. It is an adaptive one.

This is where many innovations fail. Not because the underlying ideas are flawed, but because they are not designed to operate under sustained pressure or with the necessary agility to adapt.

The Complexity of the Problem

The industrialization of frauds changes the nature of the response required.

It is no longer sufficient to detect known patterns or react to past incidents. Systems must identify behaviors that have not been seen before, anticipate emerging tactics, and operate continuously across multiple channels and geographies.

This requires more than isolated innovation. It requires systems that can learn, adapt, and scale, not once, but continuously. And behind those systems, it requires something even more fundamental: a culture capable of producing and sustaining that level of performance over time.

The way of the patent

Innovation that exists only on paper has limited impact. Innovation that survives deployment, that continues to perform as conditions change, is what ultimately defines effectiveness. In financial crime prevention, the gap between these two is critical.

In recent years, the financial sector has significantly increased its investment in AI and machine learning, with a sharp rise in patent activity across the industry. From large banks to specialized technology providers, there is a growing recognition that intellectual property can capture and formalize advances in detection, decisioning, and risk management.

According to recent data, AI-related patent filing in the financial sector grew by over 250% in the past five years: from big banks to small startups working in the space, there is a clear interest in adding the value of patents to the business.

But patents, in this context, should not be understood as an end in themselves.

They are not simply indicators of inventive capacity. They are signals of something more structural: the ability to repeatedly transform ideas into capabilities that operate reliably in real-world systems.

Strong ideas

If you obtain one patent, it suggests a strong idea. If you obtain ten, it suggests a strong team. If you obtain one hundred, it suggests a strong culture.

A culture in which ideas are not only generated, but challenged, tested, refined, and integrated into systems that must function under real-world constraints.

This distinction becomes tangible when looking at what such innovation enables in practice.

It allows financial institutions to analyze behavior across extended time horizons in real time, not only evaluating a transaction in isolation, but understanding how it relates to patterns built over weeks or months. It enables a shift from static rule-based detection to continuous behavioral modelling, improving both the precision of anomaly detection and the speed of response.

In environments where decisions must be made in milliseconds, these capabilities are not incremental improvements. They determine whether institutions can intervene while fraudulent activity is unfolding, rather than reacting after the fact.

Some of these approaches are already being deployed at scale within large financial institutions, enabling significantly faster decision execution and more robust behavioral insight across complex transaction environments.

From this perspective, patents are not about invention alone. They are about building the conditions under which innovation can endure and translate into systems that perform under pressure.

From Invention to System Performance

For an idea to matter in this context, it must pass through several layers of validation. It must be new. It must not be obvious. And it must be useful, not in theory, but in the systems that institutions rely on every day. This last dimension is often overlooked.

Usefulness, in a real-world financial system, means the ability to operate reliably at scale, under constraints, and in the presence of adversarial behavior. It means integrating into complex infrastructures, supporting decision-making in real time, and remaining robust as both legitimate usage and criminal tactics evolve.

In other words, innovation is not defined by invention. It is defined by sustained system performance.

Matching the Scale of the Threat

The industrialisation of financial crime introduces a structural asymmetry. On one side, highly organised networks operate with speed, coordination, and adaptability.

On the other hand, defensive systems have historically been fragmented, reactive, and constrained by legacy architectures. Closing this gap is not a matter of incremental improvement. It requires a shift in how systems are designed, built, and evolved.

The level of innovation required is defined by the level of organisation of the threat.

And as that threat continues to industrialize, the systems designed to counter it must do the same.

Beyond Breakthroughs

This does not diminish the importance of research. On the contrary, it reinforces it. Breakthroughs are necessary, but they are not sufficient.

What ultimately matters is the ability to translate those breakthroughs into systems that function reliably in the real world, systems that can operate continuously, adapt dynamically, and maintain performance under pressure.

In financial crime prevention, this is not an abstract challenge. It is an operational reality. And it is one that will define the effectiveness of institutions, the resilience of financial systems, and, ultimately, the level of trust those systems can sustain.

Better understand cybersecurity with the best online courses.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

The defining feature of the AI era is the speed at which AI tools have become a significant part of our working lives. Whether that’s generating content, summarizing data or automating routine tasks, AI’s speed is collapsing timelines that once took hours, days or weeks into seconds.

This is more than just a technological advancement. The rapid adoption of AI has encouraged a culture defined by instant gratification and a shift towards immediacy and expectation.

This culture shift is one of the most powerful forces shaping AI adoption, driving innovation, unlocking productivity and redefining competitive advantage.

Yet, bubbling beneath the surface are risks at a scale we’re only just beginning to understand. Employees are inputting sensitive data into AI systems, automating processes without fully understanding security implications and increasingly trusting outputs that are not properly authorized.

While organizations are increasingly confident in AI’s capabilities, the technology risks outpacing regulation and compliance. This leaves businesses vulnerable to unwarranted data risk and more cyberattacks.

The self-sustaining acceleration loop

AI is being powered by rising demands for speed and productivity. As these models become more intuitive, they remove barriers to use and are woven into everyday workflows.

That creates a feedback loop where speed becomes the priority and anything that slows it down, be it governance, security checks and/or compliance, look like obstacles rather than necessity.

At the same time, organizations are feeding these systems sensitive information with little visibility or control over where it goes, who is using it, or why. This isn’t always deliberate, but a byproduct of urgency.

We’ve seen this before. Convenience wins until the consequences catch up. From weak passwords to rushed cloud migrations, speed has often outpaced security. AI is following a similar trajectory, only faster and at a greater scale.

Regulation and compliance in catch up mode

Regulatory frameworks are also struggling to keep pace. By the time legislation is proposed, debated and implemented, the technology it aims to govern has often evolved. This leaves regulators reacting to yesterday’s risks rather than getting ahead of tomorrow’s flaws. In cybersecurity, that’s a losing game.

The gap between innovation and oversight is widening, and gaps are where threats thrive.

Cybercriminals are already using AI to scale attacks, automate reconnaissance and generate highly convincing phishing campaigns, with AI tools lowering the barrier to entry while increasing the attack surface.

As regulatory blind spots widen, threat actors won’t wait. They will move faster than the systems designed to stop them and exploit every delay.

Reframing the conversation

This is not a case against AI. Its benefits are real and, in many cases, unavoidable. The issue is imbalance, where advancement is favored and regulation and security are compromised.

We are moving too fast without the foundations to support it. As AI becomes embedded in core business processes, small gaps can scale into serious risks.

To unlock AI’s full potential without amplifying risk, we need to reframe how we think about progress. Organizations must understand their data flows in AI environments: what is used; where it goes; and how it is protected. Visibility and governance are not optional, they are the baseline.

Security must also be built in from the outset, not retrofitted. This requires alignment across technical teams, leadership and risk functions. AI cannot sit in a silo, it needs to be integrated into broader security and compliance frameworks, supported by closer collaboration between the industry and regulators.

Slowing down to move forward

To sustain momentum, we may need to slow down and create space for regulation to catch up. Building in governance, validating data use and embedding security controls will introduce friction, but it’s the kind that builds trust and resilience.

While the suggestion to slow down may feel like trying to stop a juggernaut with a stick, taking time now to pause and reflect is vital if we’re not to keep amplifying dangerous risk. A short pause now gives space to assess what is happening, what is needed which allows organizations to take back control.

Right now, AI is accelerating faster than our ability to manage the risk it creates. We need to adjust our priorities before the gap between security and speed becomes too wide to bridge.

We've ranked and reviewed the best antivirus software available.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Enterprises are deploying agentic AI at a pace that has outrun their ability to govern it.

Gartner predicts the average Fortune 500 enterprise will have over 150,000 agents in production by 2028, up from fewer than 15 in 2025.

Yet only 13% of organizations think they have the right governance in place to manage them.

The result is an execution gap: agents deployed in isolation, producing outputs nobody acts on, automating tasks rather than business processes and delivering unclear business value as a consequence.

Governance failures are an execution problem. Agents that can't interface safely with enterprise systems can't automate business processes in any meaningful way. They stay isolated helpers: producing artifacts, fielding customer queries, handling individual tasks.

The execution gap — the distance between what agentic AI promises and what it actually delivers inside the enterprise — remains largely unaddressed.

In 2026 and beyond, the guardrail problem poses an existential risk for enterprises. Adoption has outpaced controls, meaning that agentic AI is scaling faster than robust security measures can be implemented.

The speed of tech progress

The speed of tech progress can no longer stand as a rationalization for falling behind, and enterprises must address it before agentic becomes uncontrollable. Getting guardrails right will separate enterprises that realize full autonomy from those that stall out in pilots.

First, autonomy amplifies risk. Just because agentic AI can act on its own doesn't mean it requires zero human oversight. Autonomy does not equal autopilot. For agentic AI to generate real ROI, agents must do more than reason and respond. They must execute inside the business. That means interfacing directly with enterprise systems: ERP software, finance platforms, supply chain tools and the workflows that run the organization. Without that integration, agents remain one step removed from the work that actually matters.

Operational speed can compromise safety, compliance and reliability. Agents work at a blazing clip and on a more granular level than RPA. But speed becomes a moot point if agentic adoption leads to vulnerabilities such as sensitive data exposure.

Security and IT teams haven’t universally adapted to the new risk landscape. Among the risks agentic poses, "shadow AI" has emerged as a consequence of employees using unauthorized, unsanctioned AI tools or applications. When proper IT oversight or approval gets bypassed, it sets the stage for noncompliance and severe reputational damage. Departmental AI agents are proliferating without central oversight, creating security hazards and fragmented intelligence.

Governance lags far behind adoption. In this case, the guardrail gap might as well be a lack. Surveying more than 3,000 IT and business leaders worldwide, Deloitte found that just one in five enterprises reported mature governance to manage the risks of agentic AI. Autonomy without governance is a liability. This is particularly critical as we move toward the era of programmable finance, with Gartner predicting that 20% of monetary transactions will be programmable by 2030.

How to Lay the Rails Right

Agentic systems perform across a wide range of functions. When building guardrails, there must be no shortcuts. Guardrails bolted on after the fact can't account for the ways agents actually fail: corrupting data, contradicting decisions made elsewhere in the business and creating conflicts between teams acting on different outputs. Controls need to be built into how agents execute, instead of layered on top.

1. Practice measured orchestration

When enterprises accelerate AI adoption by stitching isolated tools across departments, security gaps grow harder to manage — because there’s no unified layer to anchor guardrails to. Start by scoping the broader business objective your agentic system needs to serve, not just the task.

Once you've determined what your agentic system will handle and which structured outputs will return to the workflow, built-in validation and guardrails become platform-level capabilities rather than afterthoughts bolted onto each individual agent.

2. Build governance capabilities

Without clear boundaries, agentic AI collapses. First, determine which decisions it can make independently versus those that need human approval. Real-time monitoring systems that flag anomalies and audit trails that capture the full chain of agent actions will enable accountability and continuous improvement.

3. Scale deliberately

No matter how sexy the pilot, agentic AI needs time to mature within the enterprise; you want to spot potential issues before they appear, not after. Start with lower-risk use cases and easy, single-task wins, as with fraud detection and remediation or vendor reconciliation. Avoid intricate processes with hundreds or thousands of inputs, such as the financial close of a business.

4. Guardrail gap = skills gap

While agentic AI excels at reasoning, the execution of reliable, repeatable business processes still demands deterministic systems — and human oversight to bridge the two.

To ensure smooth agentic operation in an enterprise, train your employees to move from triage, menial activities and repeated manual steps to judgment, governance and strategic decision-making roles. They absolutely require those skills. Scrum and Tiger teams can solve early problems and address early lessons, then pinpoint how agentic addresses your needs.

Putting it All Together: A Guiding Guardrail Principle

Yes, agentic AI scales productivity, but without strong guardrails, agentic AI scales risk even faster. Strategic observability and deterministic guardrails are required to ensure that non-deterministic AI stays compliant with regulatory and business standards, with reliable audit trails as well as rules for exactly when to escalate a decision or task to a human for complex exceptions or strategic oversight.

In the rush to embrace agentic, remember that the attendant tasks don’t represent a series of punch-list items. Veterans of software adoption and replacement projects know that it’s a holistic process where human actions and digital components fall into place with methodical synchrony.

Agentic AI, while it has altered the face of enterprise technology forever, rewards the same discipline every transformative technology before it has: lay the foundations carefully, and you won’t be fighting fires when it scales.

We list the best IT automation software.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

For most of the last forty years, data center performance gains came from one place: smaller transistors. Moore's Law and Dennard scaling did the work.

Each new generation of silicon delivered more performance at the same or lower power, and thermal was a maintenance problem, not a performance limiter.

Cooling sat in the background. Operators measured it through PUE, optimized for it where convenient, and otherwise treated it as overhead.

That world is over.

Dennard scaling broke years ago, transistor efficiency gains are leveling off, and AI accelerator TDPs have climbed from 700 watts in the H100 generation to over 1,400 watts in current Blackwell deployments, with NVIDIA's upcoming Rubin platform expected to push further.

Thermal is no longer something that happens after the architectural decisions. It is now the binding constraint on how much performance a chip can sustain, and it is becoming one of the most strategic choices an AI data center operator can make.

Why this matters now

The macro numbers explain why this matters now. Data centers already consume up to 4.5 percent of total U.S. electricity production, a figure projected to reach 12 percent by 2028. McKinsey estimates global data center spending could approach $7 trillion by 2030, and that data center power demand will reach 220 gigawatts in the same window.

None of that capacity arrives quickly. New transmission lines and substations now take five to ten years to permit and build, which means operators cannot simply order more power when they need to scale.

The result is a hard pressure to extract maximum performance from the power they already have under contract. That pressure is what is reshaping how the industry thinks about cooling.

Cooling is no longer just an afterthought

For years, cooling was measured as an efficiency loss, captured through metrics like Power Usage Effectiveness (PUE) that quantified how much energy was burned on overhead before reaching the IT load. Today, the more meaningful metric is how much useful compute you extract per unit of power. NVIDIA's Jensen Huang now describes this as "performance per watt" or "tokens per watt" for AI workloads, and cooling plays a direct role in both halves of that equation.

Direct-to-chip liquid cooling has become the new baseline because it removes heat far more effectively than air. But even direct-to-chip is being pushed to its limit by 1,000+ watt accelerators, and most current deployments still require facility water around 30 degrees Celsius to stay within ASHRAE W2 and W3 envelopes, which means chillers running for much of the year in warm climates.

Better thermal management has effects on both sides of the tokens-per-watt equation. It reduces facility overhead, so more of the contracted power reaches the rack. And it allows chips to operate closer to their full thermal headroom, sustaining higher performance for longer.

Those gains compound. Recent UCLA study has shown that combining a 17 percent improvement in facility efficiency with a 15 percent gain in server-level performance per watt from better thermal management translates to roughly 35 percent more tokens per watt within the same power envelope. In a 10 megawatt facility, that is more than a megawatt of additional usable compute, with no additional grid commitment.

At GTC 2026, NVIDIA CEO Jensen Huang made this argument explicitly. He told the audience that beyond the silicon roadmap, infrastructure-level optimization across power and cooling represents another factor of two in performance still on the table. "There's no question in my mind there's a factor of two in here, and a factor of two at the scale we're talking about is gigantic," he said.

That gain does not come from a smaller transistor. It comes from rethinking how power and thermal energy move through the rack. Recent UCLA study suggests that at least one third of that infrastructure-level gain is attributable specifically to cooling. Cooling is no longer a support function. It is a primary lever for performance.

Water is becoming a hard constraint

Power is not the only pressure point. Water is emerging as an equally critical and often more immediate constraint on data center expansion. Traditional cooling architectures often rely on evaporative processes that consume vast amounts of water. According to the Environmental and Energy Study Institute, large data centers may use up to 5 million gallons per day, comparable to the daily water use of a town of 10,000 to 50,000 people.

This is drawing notice from regulators and communities in already water-stressed areas. The result is longer permitting cycles, higher project risk, and in some cases new developments paused entirely. States and municipalities are also implementing stricter reporting requirements and adjusting electricity rate structures specifically for data centers.

Operators now have to factor water alongside power into site selection. Facilities that minimize energy waste and reduce or eliminate water consumption are better positioned to navigate this environment.

The shift toward next-generation cooling

In response, the industry is entering a new phase of cooling innovation. Air cooling is no longer sufficient for high-density AI workloads. Liquid cooling has become the baseline, but within liquid cooling, not all approaches deliver the same efficiency or scalability.

The next wave of innovation focuses on improving heat transfer at the source: removing thermal energy more effectively at the chip level while reducing system-wide overhead. Some of these approaches draw on heat transfer techniques refined in other high-density power industries such as nuclear power generation, where the challenge of moving large amounts of thermal energy from a constrained physical space has been studied for decades.

The goal is straightforward. Better cooling enables higher rack densities, allows operation at higher facility water temperatures, and reduces or eliminates reliance on water-intensive heat rejection. Just as importantly, the next generation of cooling architectures is being designed to integrate with existing data center footprints, so operators can evolve their infrastructure rather than rebuild it from scratch.

NVIDIA's Vera Rubin platform, announced at CES 2026, was a clear signal of where this is heading. Vera Rubin is designed for 45 degree Celsius supply water, which means dry coolers can do most of the heat rejection year-round and mechanical chillers become optional in most climates. That is a fundamental shift in how cooling infrastructure will be designed for the next decade.

A defining moment for data center design

The data center industry is at an inflection point. AI compute demand is accelerating, and every resource needed to support it, power, water, physical space, is becoming harder to secure. Cooling sits at the intersection of all three.

It determines how efficiently power is used, how much water is consumed, and ultimately, where infrastructure can be deployed. The operators that recognize this now will have a sustained advantage. How to keep data centers cool under AI workload pressure has become one of the most strategic decisions in modern infrastructure.

We feature the best web hosting services: tested and reviewed.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Generative AI has moved quickly from experimentation into early production use in many enterprises. However, very few can confidently forecast what it’s going to cost them in six months.

For a technology that has consumed so much board-level attention and capital, that reflects a lack of certainty, and one that some technology leaders may privately recognize as true of their own organizations.

The spend is real and the direction is clear, but the number at the end of the year can remain genuinely uncertain.

To capture a glimmer of the confidence driving the infrastructure race, Amazon’s CEO has indicated it expects to spend heavily on IT infrastructure to support AI, with an estimated $200 billion in AI capital spending, arguing it is “not going to be conservative” in how it invests in the tech.

In practice, what makes AI different from the infrastructure investments that came before it is not the scale of the commitment but the nature of the consumption.

Cloud computing was unpredictable when it arrived too, but it eventually settled into patterns that finance teams could learn to model. AI hasn’t settled in the same way yet, and much of the reason comes down to how it is being used.

A great deal of enterprise AI use remains exploratory, which is part of what makes forecasting harder. And unlike cloud, which stayed largely within technical teams for years before spreading, AI is moving across the whole organisation almost immediately. That changes everything about how you try to govern it.

The limits of financial visibility

On the surface, some forms of AI appear to offer what earlier infrastructure lacked: clean, granular, real-time data and what it costs. But across the rapidly growing landscape of technology providers leveraging AI in some way, many do not.

In some cases, token-based pricing is precise in a way that early cloud billing never was, and for finance teams accustomed to working with far less, it can feel like a step in the right direction for solving the visibility problem.

We unfortunately still have a long way to go, since simply understanding what was spent last month tells you very little about what will be spent next quarter, particularly once adoption moves beyond the teams who originally shaped the business case.

One must consider that teams across legal, HR, and customer operations are not thinking about token economics (tokenomics). They’re only thinking about whether the tool works.

Cost exposure builds not through any single decision but through dozens of small expansions, each reason in isolation, none of them reflected in a comprehensive forecast. By the time anyone joins the dots, the demand curve has already moved.

Extending the disciplines that already exist

The organizations who are doing a better job managing AI spend have tenured experience managing consumption-based technology. IT asset management (ITAM) teams for example often have more experience dealing with more fixed constructs like users or seats, which makes the consumption-based nature of AI far more challenging.

FinOps teams on the other hand have grounded experience in managing consumption that originated in public cloud. FinOps teams may therefore better positioned to deal with the new tsunami of AI consumption and spending, ensuring that it is governed as adoption scales.

FinOps has also been broadening its scope beyond the initial roots in public cloud, with AI cost management now sitting firmly within that remit for many, a shift reflected in how the FinOps Foundation is increasingly incorporating AI into its guidance. Part of that expansion is about forecasting demand that behaves differently from conventional workloads.

There is also growing interest in whether AI itself can support FinOps practices, particularly in anomaly detection, optimization and, over time, forecasting, as consumption patterns become harder to model.

The challenge is applying FinOps practices early enough so that governance shapes how AI scales, rather than scrambling to restore control once spend has already outpaced oversight.

The compounding difficulty of legacy environments

For organizations whose technology estates were built around consistency, extending governance into AI is harder than it sounds.

AI-first organizations design with cost in mind from the beginning, treating inference the way they would any other product input, with economic constraints shaping architecture decisions before commitments are made.

Retrofitting AI into legacy infrastructure means something different. Existing commercial commitments and operating models do not adapt quickly to a consumption model that is inherently variable, and that friction has a direct bearing on cost.

The difficulty is often that AI is being introduced into environments built around very different assumptions about how demand behaves, and that is part of what makes forecasting harder.

The challenge is not simply new spend, but expenditure ballooning in environments where oversight and control are already difficult to maintain.

Organizations navigating this will tend to run controlled experiments before broad rollout and are deliberate about how adoption spreads. In practice, that is often about containing unmanaged adoption early, before usage patterns, costs and dependencies become harder to unwind.

That same exposure increasingly carries beyond internal governance. As AI appears more often in customer procurement conversations, questions that were once largely internal are starting to be probed externally too.

For organizations whose governance has not kept pace, those questions can force a level of clarity they may not yet be prepared to provide.

From activity metrics to business outcomes

Beyond governance and cost control, there remains a harder question, which is whether AI investment is producing meaningful business value. Most leadership teams are not yet in a position to answer that with confidence, and the metrics currently reaching the board are not making it easier.

Model usage, inference volumes and compute consumed describe activity without explaining value. It is easy to build a compelling board update from consumption data without addressing whether any of it is moving the business.

What gets closer to an answer is understanding whether individual inferences are delivering something a customer would pay for, or something that meaningfully reduces cost or risk.

Incremental business outcome per pound or dollar of AI spend is a harder measure to produce, but it is closer to the economics that matter because it requires a clearer position on what AI is actually delivering.

That is precisely where many organizations are still finding the work harder than it looks, particularly as AI deployment moves ahead of the models used to understand cost and value.

That disconnect matters more as the market expands, because where those economics remain unclear, cost exposure can build in ways that are harder to recognize early and harder to contain later.

For many enterprises, the challenge ahead is scaling AI without allowing spend to outrun the value it is meant to create.

The best cloud storage: tested, reviewed and rated by experts.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Agentic AI is moving rapidly from boardroom ambition to enterprise reality.

Gartner forecasts that roughly 40% of enterprise applications will incorporate task-specific AI agents this year, up from just 5% last year.

This surge forces every CIO, CISO, and technology leader to consider: What should AI be allowed to access, and how should it operate once inside the enterprise?

Many organizations begin by embedding AI agents directly into legacy systems, connecting them to backend databases, APIs, and workflows in the name of speed.

While this inline approach can work in modern, well-governed environments, it often bypasses the approval workflows and controls that legacy systems were built around. Agents can access restricted data, skip approvals, or execute transactions without a complete, attributable record.

The result is a growing governance gap. Decisions tied to sensitive data can’t be reliably reconstructed or defended with the same confidence as human-driven work. Even advanced models stall in pilots because organizations can’t prove how outcomes were produced.

The solution is not to slow AI adoption. It’s to change how AI interacts with the systems that already run the business.

When AI bypasses the system, it breaks it

Consider a finance workflow in an ERP software system. An agent updates vendor bank details and pushes a payment through a fast-track path, bypassing a required approval step and segregation-of-duties check. Later, when the transaction is questioned, the organization can’t prove who approved the change, why it was made, or whether proper controls were followed.

That’s where accountability breaks down. Changes are made inside core systems, but the evidence is incomplete, inconsistent, or disconnected from the system of record.

Emulated human behavior offers a more secure and practical path. These agents operate exactly as a human employee would: logging in with standard credentials, navigating the existing user interface, reading screens in context, following established workflows, and executing tasks while remaining fully subject to every control already in place.

No new APIs. No raw backend data exposure. No rewriting of decades-old business logic or security rules. The guardrails designed to protect against human error or misuse — validations, permissions, approvals, and audit logging — remain 100% intact.

This UI-first approach is especially effective for organizations running mission-critical processes on older platforms. Building secure, governed APIs for legacy systems is expensive and time-consuming, often leaving out protections built into the interface layer.

While emulated human agents may not match the speed of direct backend calls, they provide far more valuable enterprise advantages: immediate deployability, ironclad accountability, and zero disruption to proven controls. Secure operation doesn’t require avoiding AI. It requires rethinking how it fits into the systems around it.

Preparing for emulated human in the enterprise

Three priorities can help organizations prepare for the emulated human approach as AI scales into critical workflows.

1. Place AI at the points where work happens

Most enterprise AI strategies assume deeper backend integration creates better automation. In environments shaped by legacy systems, it often does the opposite: introducing new complexity while bypassing the workflows and controls already built into the interface layer.

Instead, focus AI at the points where it can operate without requiring systems to be rebuilt. This approach dramatically reduces integration overhead, limits exposure of core systems, and allows AI to scale within existing operating models rather than forcing costly modernization.

2. Align AI accountability with human accountability

Agents should operate under named identities and the same policies as employees. They preserve approval workflows, follow role-based permissions, and generate the same audit artifacts — including log entries, change histories, tickets, and recorded approvals — that organizations already rely on to review human activity.

This removes the dangerous two-tier governance model where AI operates under different standards than employees. Organizations can maintain visibility, accountability, and established compliance and risk management controls as AI takes on greater responsibility.

3. Design for adaptability rather than brittle automation

Traditional robotic process automation (RPA) relied on rigid, click-by-click scripts that broke the moment screens changed or exceptions appeared. Emulated human agents interpret context in real time, adjust to variation, and continue operating, just as skilled employees do.

That adaptability is essential in dynamic enterprise environments where policies change, exceptions are common, and systems are rarely static. Instead of constant break/fix maintenance, organizations gain AI that can operate more resiliently inside real-world workflows.

Scaling AI with the systems already in place

As agentic AI scales, enterprises will be judged not only by the intelligence of their systems but by their ability to govern them. The pressure to balance innovation with control will only intensify.

The most durable strategies will be those that embed AI safely within the systems already in place, rather than racing around them. When an agent’s actions can be audited and justified with the same rigor applied to a human colleague, it’s finally ready for production.

That’s how secure, scalable AI will be defined in the enterprise.

We feature the best small business software.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Every small business is looking for an edge.

For some, that means protecting margins. For others, it means finding new customers, sourcing better products, entering new markets or simply running the business with less friction.

More often, it means trying to do all of these things at once. Knowing which opportunity deserves attention first and having the time and resources to act on it has always been a challenge.

For decades, this has forced smaller businesses into a trade-off: choose one priority and hope this is the right decision or waste all your time and money and miss an opportunity elsewhere.

The era of agentic business changes this. According to the British Chambers of Commerce (BCC), more than half of UK SMEs (54%) are now actively using AI – an increase from 35% in 2025.

In recent times, the technology has moved from passive assistance - writing, summarizing and answering questions – to autonomous execution.

AI is therefore no longer limited to responding only when prompted but can operate continuously in the background.

A wide spectrum of business functions

In this way, AI agents can now handle a wide spectrum of business functions end-to-end. This includes building digital storefronts, writing product listings, offering dynamic pricing, providing customer service, generating market research, and supporting with marketing.

The real value lies not in overnight task automation but better decision-making with less friction and fewer blind spots. In fact, according to the same BCC research, SMEs using AI report strong net productivity improvement expectations (+71%), while those planning to adopt or unsure show far lower optimism. Importantly, agentic AI is levelling the playing field with larger companies.

The latter have historically had an advantage because they can afford dedicated teams for each business function. SMEs, on the other hand, have had to rely on lean teams, founder instinct and whatever time was left after the urgent work was done. Within agentic businesses, there is immediate access to capabilities which once required high headcount or expensive IT systems.

Examples include testing a new product category, launching into a new market or trialing a marketing campaign with far less operational risk than before. Rather than spending weeks gathering information manually or coordinating across multiple systems, AI agents can help businesses identify opportunities and execute tasks in real time.

Crucial for smaller businesses

In addition, agentic AI has been crucial for smaller businesses looking to grow internationally. It can help them localize product listings and marketing content for different markets, coordinate supplier communications across time zones, and analyze regional demand trends in real time. This reduces much of the operational complexity traditionally associated with cross-border trade and gives SMEs greater confidence to explore new markets that may previously have felt out of reach.

So, the SME conversation around AI needs to move beyond productivity. Saving time matters, but it is not the full story. The bigger opportunity lies in performance: simplifying complexity, reducing avoidable risk, helping businesses act on information earlier.

For an SME, one missed supplier issue, one misread market signal or one poorly timed product decision can have an outsized impact. Becoming an agentic business helps reduce that exposure with complex information easier to monitor, compare and act on. However, it does not remove the need for human judgement. In fact, it raises the value of that judgement by giving business owners clearer options and more time to focus on strategy.

The most successful uses of AI will not be the most futuristic but the most useful, offering practical, transparent information, built around real commercial pain points. The first wave of AI helped SMEs create faster, but the next wave will help them operate smarter.

For SMEs, the question is no longer whether AI can help. It is how quickly they can put AI agents to work on the decisions that determine how they compete, grow and scale.

70+ of the best AI tools tested and reviewed.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

The American entrepreneur Palmer Luckey has played a major part in many major names in the tech industry, including founding Oculus VR, but has risen to prominence for his stewardship of Anduril Industries in 2017 – which puts artificial intelligence (AI) at the heart of its operations and capabilities.

The question of AI ethics

Luckey has long been a proponent of the use of AI in defense, with the Anduril founder even naming his company as such because it shares the acronym.

Speaking with Fox News Sunday in December 2025, he outlined a philosophy that it's much safer, in ways, to incorporate the best possible technologies into military capabilities than it is to ignore emerging innovation, whether AI or quantum, and be left with 'lesser' tech.

For Luckey, there's seemingly no point in taking the moral high ground when talking about matters of life and death – when the very nature of the business is morally called into question to begin with.

The future of military technology

Luckey's statement challenges the common orthodoxy that humans must always be in the loop over high-stakes decision-making, especially with matters of life and death. And, for that reason, it's highly controversial.

The direction of travel is not, however, up for debate, with plenty of examples of AI becoming increasingly prominent in international conflicts.

Ukrainian officials, for example, recently revealed that in 2024, the nation's military used drones to kill Russian soldiers – marking the earliest reported example of the autonomous killing of humans.

Without doubt, it won't have been the last such incident, and points to a future in which those engaged in conflict are likely to use every tool at their disposal to achieve their military aims, regardless of the moral implications.

Retail fraud used to be relatively straightforward.

A stolen card, a fake account, or a suspicious transaction pattern that could be flagged and blocked before serious damage was done.

That version of fraud is still present, but it is no longer the main problem.

What’s appearing now is something more coordinated, automated, and harder to detect in real time: AI-powered fraud rings that behave less like individual bad actors and more like distributed systems.

They test, adapt, and scale in more sophisticated ways that increasingly mirror the technologies retailers themselves are adopting.

Fraud is no longer just responding to ecommerce; it’s evolving alongside it.

From isolated fraud to coordinated systems

For years, fraud prevention has largely focused on individual events: a suspicious login, a stolen card attempt, a bot probing checkout flows. But that model is breaking down.

What is now emerging is coordinated fraud activity that behaves more like a network than a series of isolated incidents. These groups combine automation, synthetic identities, and increasingly realistic AI-generated content to simulate genuine customer behavior at scale. The result is not only more fraud, but fraud that blends into normal digital traffic.

At the same time, fraud rings are executing high-velocity attacks that look more like engineered systems than opportunistic crime. One recent example involved an estimated $4.2 million in fraudulent activity over 48 hours, driven by synthetic identities, spoofed devices, and rapid transaction flows reaching around 180 per minute.

What is notable is not only the scale, but the structure. These are not isolated attempts. They are coordinated operations designed for speed, repetition, and adaptation.

AI is lowering the barrier to fraud

The most important shift is accessibility, as well as scale. Generative AI has significantly lowered the barrier to entry for fraud. Tasks that once required technical expertise or coordinated effort can now be executed using widely available AI tools.

Fraudsters can generate synthetic identities, fabricate supporting documents, and create convincing digital artefacts that simulate legitimate customer behavior in a matter of minutes. This includes everything from account creation to behavioral patterns across browsing, purchasing, and post-purchase interactions.

One of the clearest examples is the rise in returns abuse, which has increased by 15% in the past six months, largely driven by the ease and scalability of AI-doctored images.

In practice, this means fraudsters can submit highly realistic images of cracked, damaged, even moldy, or malfunctioning goods. These images are often convincing enough to pass initial review processes, particularly when combined with legitimate order histories or stolen account credentials.

In several documented cases, fraud rings have used newly created accounts to purchase low-cost goods, then submitted AI-generated images to claim refunds without returning the product. In some instances, empty boxes are shipped back instead, all while refunds are still processed.

Another coordinated operation targeting home goods and fashion retailers used a similar approach at scale, generating approximately $800,000 in fraudulent refunds through repeated low-value claims designed to avoid detection thresholds.

Individually, these cases may appear manageable. But collectively, they show a clear trend: fraud is increasing in sophistication and scale.

The shift toward agent-driven commerce

The next phase of this evolution is already on its way, and it’s closely linked to the rise of autonomous digital agents.

Over the second half of 2025, agentic activity surged by over 2000%. While much of this growth is tied to legitimate use cases such as shopping assistants and automated customer workflows, it also introduces a new layer of complexity for fraud detection. The same systems that allow agents to complete purchases on behalf of consumers can also be manipulated to automate fraud on a larger scale.

Instead of manually creating accounts or placing orders, fraudsters can now orchestrate entire attack chains using agent-based automation. This changes the nature of fraud from individual actions to continuous, self-executing systems. This matters because it shifts the detection problem. When fraud resembles legitimate automation, the distinction between genuine and malicious activity becomes harder to define using traditional rules.

At the same time, retailers are already seeing fraud patterns adapt to this environment. Attackers are increasingly mimicking normal customer journeys, spreading activity across devices, accounts, and timeframes to avoid detection. The result is a system where fraud does not look like fraud until after the fact.

Why traditional fraud tools are falling behind

Most existing fraud detection infrastructure was not designed for the current conditions as they stand. They rely heavily on static rules, historical patterns, and known indicators of malicious activity. But AI-led fraud doesn’t necessarily follow predictable patterns. It adapts in real time, varies its behavior based on changes in the attack surface, and can scale in ways that overwhelm rule-based systems.

Even machine learning models trained on historical fraud data struggle when faced with synthetic behavior that has no direct precedent. This creates a widening gap between how fraud actually operates and how it is detected.

Consequently, many retailers are forced into reactive positions, identifying fraud after fraudulent transactions have already been completed rather than preventing it in real time. This is particularly challenging in areas like returns and refunds, where fraud is often indistinguishable from legitimate customer claims at the point of interaction. The core issue lies in timing alongside detection accuracy.

What comes next for digital trust

The trajectory of fraud is closely tied to the progression of ecommerce itself. As AI agents take on a larger role in how consumers find, compare, and buy products, retailers face a more complex question than simply whether a transaction is legitimate.

They need to determine who, or what, is actually behind the transaction. Is it a real customer? A legitimate AI assistant acting on their behalf? Or a synthetic system designed to imitate both?

The challenge now is no longer just detection, but judgment in real time. Because in an environment shaped by AI on both sides of the transaction, risk and verification can no longer sit at a single point in the process. They must be continuously reassessed throughout the customer journey.

Manage customers for a new business more effectively with the best CRM for startups.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

AI agents are reshaping how enterprises automate work, but their effectiveness depends on access to sensitive systems and data.

The paradox is that granting them the permissions they want creates new attack surfaces that organizations aren’t yet equipped to handle.

This is the defining tension of the AI era.

AI agents are proliferating across enterprises with 91% of organizations already using them yet only 10% have a clear IT management strategy in place.

This gap matters because as these systems grow more autonomous and more deeply embedded in workflows, enterprises are operating without clear visibility, meaningful oversight and control over how their AI agents behave.

The access problem

Our recent research revealed how agents running on OpenClaw, an open-source AI agent automation platform, could expose credentials and leak sensitive information when attackers compromised the communication channels controlling them.

To appreciate the scale of this risk, we must first understand the platform itself. OpenClaw combines a chatbot-style interface with access to external tools and large language models.

Users can then configure agents to browse the web, read and write files, manage inboxes, execute commands, or interact with other machines. In many cases, they’re designed to operate autonomously with minimal human oversight.

That level of access is what makes agents powerful, helping many to manage everyday admin and time-consuming tasks. However, this power is a double edged-sword and can make them a risk to businesses.

When agents become attack surfaces

Agents need access to tools, accounts, applications, the web and more to be useful. Often, this means an agent needs access to secrets: API keys, personal access tokens, credentials, .env files, OAuth tokens.

The agents/models are by default prompted to be as helpful as possible, and that characteristic starts to pose some particular concerns when it comes to credentials and tokens. If an agent such as OpenClaw can’t access a resource, it will ask for credentials right in the chat, exposing those secrets within the context window. Agents will happily store API keys in their unencrypted configuration files, which information-stealing malware is starting to target.

Remote access capabilities could effectively create a back door into enterprise environments. If an attacker gained access to the communication channel controlling an agent, such as a messaging or remote access platform, they could potentially gain access to everything the agent itself could access. In an enterprise context, this is a nightmare.

The paradox of recognized risk

Perhaps the most revealing finding was that some agents recognize risky behavior while simultaneously carrying it out. This underlines how their decision-making ability and autonomous operations can be a business risk.

In one test, an agent correctly identified that exposing an OAuth refresh token through an unencrypted communication channel represented a serious security violation. But it then proceeded to share the token anyway before expressing concern about its own decision.

Organizations should not rely on the invisible guardrails that frontier model providers put around agents. They’re easily circumvented.

But an AI agent cannot divulge credentials that it doesn’t have access to. This is why the conversation around AI agent security cannot focus solely on stronger guardrails. Attackers are already finding ways to manipulate agent behavior through prompt injection, social engineering, and compromised communication channels.

Governance, not just guardrails

AI agents are essentially identities within enterprise systems and need to be managed as such. They perform actions and make operational decisions in ways that increasingly resemble human employees or privileged service accounts. Yet many organizations are deploying these systems without applying the same governance standards.

Most businesses already understand the importance of least-privilege access, audit logging, identity management, and access reviews for employees. AI agents should be subject to the same principles. That means limiting what agents can access, avoiding long-lived credentials wherever possible, and ensuring sensitive information is stored securely through centralized systems with human oversight.

Organizations also need visibility into where agents are deployed, what tools they can interact with, and how to disable them quickly if something goes wrong. If an agent goes rogue, there needs to be a “kill switch,” a way to immediately revoke an agent’s access to resources and shut it down.

Agentic AI systems could deliver major operational upsides, but deploying them without robust identity and access governance introduces significant security risk. As these systems become more deeply embedded across enterprise environments, organizations must stop treating them as experimental tools and start governing them as part of the digital workforce.

This means managing the full lifecycle of agents, from knowing which agents are deployed, what resources they access to and keeping a full audit trail so no one can say, “I don’t know what happened. The agent did it.”

There’s no reason why conventional security wisdom, such as the principle of least privilege, lifecycle management and robust logging, should be thrown out in an agentic age. In fact, it’s more relevant than ever.

We've tested and reviewed the best cloud storage.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Scaling a billion-dollar company in Europe has historically been more difficult than it should be. Not because there aren't enough ambitious founders; rather, it’s because the conditions to scale, regulatory and funding constraints to name but two, have never fully matched the ambition.

However, the game is changing. AI is rapidly making those ‘constraints’ less relevant and with the rise of multi-agent systems, startups can operate with the capability of a larger organization.

The question now is no longer whether Europe can produce the next wave of ‘soonicorns’, (startups nearing a $1 billion valuation), but whether decision-makers are willing to abandon outdated models and build for an AI-native future.

If they do, we will soon see a new foundation for European startups, one where agility, enterprise-grade governance and AI-native architecture are baked-in from the outset.

The architectural opportunity

Europe is not short of successful startups. However, for the region to continue producing top players, we will need to see proactive change from companies - adding intelligent AI features to an existing process or product simply won’t be enough. It involves rebuilding a company’s organizational structure, something that’s only possible by multi-agent systems.

Startups no longer need to wait until they have the necessary resources to take on complicated operations. Instead, they can break those issues down into specific, identifiable problems and assign specialized AI agents to tackle them. These agents will be coordinated, efficient, and able to operate at a speed that is incomparable to a human team.

The knock-on effects are huge. Product cycles shrink and teams can concentrate their efforts on tasks that genuinely call for human judgement. Additionally, technical expertise is no longer restricted to well-funded teams, since vibe coding speeds up prototyping and AI lowers the barrier to building advanced systems. For Europe to take the lead on AI, it must start with its data foundations.

The database problem nobody is talking about

AI strategy is often the main topic of conversation in boardrooms across Europe – which models are appropriate, what use cases should be prioritized and which teams to hire. Infrastructure is often, mistakenly, absent from that discussion. In particular, the database infrastructure that the majority of companies are still developing and why it can be subtly hampering the startups with the greatest potential.

Traditional databases were designed for the past, built for slow applications, fixed infrastructure and data that is handled by humans. This worked well until AI agents came into the conversation.

AI agents require quick, dependable and instantaneous data access to perform real-time, complex actions. For startups attempting to scale quickly, building on the correct foundations is the difference between stalling and success.

The infrastructure conversation is finally catching up, with a concept known as lakebase designed to support this transition. It delivers the reliability of an operational database and the openness of a data lake in one centralized place, so teams can run transactions and analytics without juggling systems.

It enables fast access to data, scales easily through separated storage and compute, and fits modern development habits like instant branching and versioning. A lakebase gives founders an edge that previous startups never had; the opportunity for both their developers and AI agents to build, test, and ship applications quickly, without the constraints of old online transaction processing (OLTP) setups.

What founders must change now

With AI progressing at an unexplainable rate and investors asking questions, a knee-jerk reaction is usually to add a new AI capability to what currently exists. However, this is a short-term solution to a long-term problem.

Startups that approach AI as structural rather than an add-on will define what comes next. This involves raising challenging issues early on, such as how this business should be built if AI is doing a large amount of the work, rather than just what AI is capable of achieving for this product.

Instead of figuring it out after the company has scaled, founders are forced to make architectural decisions early, establishing what agents own and how they work together to make sure humans are kept in the loop.

It also means realising that governance and speed work in tandem. The entrepreneurs who incorporate guardrails early enough that they never become a barrier are the ones who grow the fastest. Integrated into the design from the start, enterprise-grade governance is a hidden competitive advantage. Instead of having to rush to catch up later, it enables you to scale with certainty.

Build it right or build it twice

The time has arrived for Europe, but it won't wait for businesses to continue bolting AI onto infrastructures that weren't designed to support it.

Soonicorns won't be determined by how much they raised, how many AI tools they have used or, how fast they delivered. They will be characterized by the caliber of the decisions made about AI-native architecture, AI agents, and whether or not humans are involved.

The tools are there and the market is shifting. Whether European founders are prepared to build with the same ambition they offer is the only true question that remains. When leaders build the appropriate unified data foundations, everything else will fall into place.

We've featured the best AI website builder.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Most companies don’t understand that today’s AI tools are capable of fundamentally transforming how daily knowledge work is done.

This is because they’re using AI in an unsophisticated way and aiming it at the wrong place.

But this level of transformation is already happening, as millions of knowledge workers have figured out, and as enlightened companies are starting to recognize.

To delve deeper, you first need to understand that most knowledge work is invisible. The essence of knowledge work—thinking, processing, judging, ruminating, planning, mulling—happens in workers’ heads, unseen.

Unfortunately, workplace AI is currently deployed into the knowledge systems that are visible, the outputs —emails, documents, chats, meetings, etc. It doesn’t matter how good the AI is, because when it operates at this level, it’s too late to really transform how the work is done.

To give a practical example, when you need to create a project deliverable, 80% of your effort is likely spent creating the first draft, with the remaining 20% polishing into a final deliverable. AI workplace assistants do a great job with that final polish (which we like, thank you). But to truly transform how work is done, you need AI to help with the underlying, unseen 80% effort used to create the first draft.

The real opportunity: A practical model for AI-driven work

The good news is AI is fully capable to help transform that 80%. This does not require waiting for “better” models or AGI. All you need to do is change how you’re using AI today, by integrating existing LLM-based tools into that invisible thinking portion of your work, rather than just keeping it at the surface-level work outputs.

While the AI vendors haven’t exactly made this intuitive (yet), using AI in this way has exploded in popularity since the beginning of 2026. In practice, the basic approach is to use AI the way software developers do, not as a one-off tool but as something that builds context over time.

Move beyond web-based interfaces where every conversation restarts from scratch

Create a centralized repository, put your critical files into a folder on the business computer which you give AI access to. Start with the classic things (deliverables, meeting notes, project plans, etc.)

Before doing any work, ask AI to interview you about your work style, what’s important to you and your personal preferences.

Review and refine AI’s understanding, ask it to scan through all your files to synthesize your latest thinking, ideas, story arcs, writing style and any other “intelligence” it can determine from your work. Review its findings and go back-and-forth until you feel it has a good understanding of you, your work and your style.

Build upon each session. A crucial step is having the AI tool understand this is not a one-time or manual exercise. Instead, a continued process to create, maintain, organize and update the files, based on what it learns about you over time, each subsequent AI session builds on all the work you’ve done together and what it has learned about you.

In essence, you are asking your AI to create a personal Wikipedia-style repository which gives your AI system an ever-growing continuous context library perfectly built and tuned just for you and your work.

Using AI like this doesn't require a new product or company, but a new way to leverage current tools. This is often called a “second brain”, “AI context vault”, “LLM-powered personal Wiki”, or something similar, and you can do this with any LLM vendor or product.

Most AI vendors now allow users to connect their LLM platforms into other business systems (like email, chat, document stores or productivity suites), which lets workers connect their personal knowledge systems into corporate apps and data.

Workers who use AI in this new way report fundamental shifts in the way they work within the first few hours. After a few days, many workers declare they will never go back to the “old way” of working again.

The tradeoffs to consider

Using AI to transform work in this way is not without its downsides, especially from the corporate perspective.

First, all the “classic” security complexities still apply: How do you know the AI did what it said it was going to do? How do you know it didn’t hallucinate? How do you trust it won’t spin out of control and email all your contacts with nonsense?

Addressing this involves many of the things you probably know but haven’t taken time to investigate yet, including configuring alternate accounts with restricted permissions for AI or setting clear guidelines for when and how AI-generated outputs will be reviewed.

This new process also requires asking workers to slow down and verify what their AI generates, which is pretty much the opposite of why they started using AI in the first place.

Another challenge is visibility. Much of the “back-and-forth” work - which previously happened in the open - now happens within the AI tool and the worker’s personal context vault, where it’s less visible to coworkers and management scrutiny. Individual workers view that as a positive, but to organizations, it can be a liability.

Lastly, when workers build personal AI context vaults using their personal AI subscriptions, the company can’t prevent the worker from taking all that context with them when they leave the company. Companies need to buy proper enterprise AI subscriptions which they can link to corporate SSO and DLP systems. The downside is that enterprise AI pricing is completely different from consumer pricing, and workers using AI like this via enterprise systems can easily consume thousands of dollars of tokens per month.

The bottom line is that today’s AI can fundamentally transform work, but only if there is a mindset reset around how it is being used.

This new approach introduces added complexity. Organizations will need to spend more time understanding, managing, and securing AI differently, but it’s clear that AI operating in this way is inevitable, so the time to start thinking about AI as a “second brain” is now.

We feature the best free office software.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

The UK’s autonomous vehicle (AV) sector is entering a period of rapid acceleration. With London preparing for the rollout of driverless taxi services later this year, and regulatory backing strengthened by the Automated Vehicles Act, the shift from experimentation to deployment is becoming tangible.

That momentum is already visible on the capital’s streets. Waymo is currently testing its autonomous ride-hailing service in London, navigating complex urban environments ahead of its planned commercial launch. But as physical deployment accelerates, a more fundamental bottleneck is emerging.

The central challenge is no longer whether autonomous vehicles can navigate roads, but whether the industry can consistently demonstrate that they are making safe, compliant decisions in real-world conditions.

Without that capability, progress toward higher levels of autonomy will stall, regardless of how advanced the underlying driving systems become.

The industry’s hidden bottleneck

Recent incidents in London illustrate the challenge. Reports of an AV entering a taped-off crime scene in Harlesden, or repeatedly turning into a Shoreditch no-through road, highlight how unpredictable dynamic urban environments remain for automated systems. Modern AV systems already perform well at perception.

Using combinations of cameras, LiDAR, radar and AI models, vehicles can detect lanes, pedestrians and hazards with increasing accuracy, and AV companies have now logged tens of millions of autonomous miles globally.

However, the real challenge lies in the transition to Level 4 autonomy, where legal liability shifts from the human driver to the manufacturer. To secure regulatory approval and public trust, companies must be able to explain exactly why a system behaved the way it did in ambiguous situations, such as navigating a temporary road layout, conflicting signals, or unusual pedestrian behavior.

This is where current machine learning approaches fall short. While effective at pattern recognition, they typically operate as “black boxes,” offering limited insight into how individual decisions are reached. In a safety-critical sector like automotive, this lack of transparency creates a major commercial and regulatory constraint.

Manufacturers and regulators need definitive evidence that systems are acting in accordance with local road rules before they can deploy at scale.

The missing layer in autonomous intelligence

To bridge this gap, the industry is increasingly turning to knowledge-based AI, an alternative to large language models that uses carefully curated expert knowledge and structured reasoning to correctly answer complex, high-stakes questions.

Unlike purely data-driven models that infer behavior statistically from past training data, knowledge-based systems combine sensor inputs with explicitly defined rules, traffic laws and domain expertise. Rather than relying solely on probability, they enable vehicles to reason through decisions using structured logic.

That distinction is critical in autonomous driving, where edge cases are difficult to predict and regulatory scrutiny is high. While machine learning remains essential for perception and pattern recognition, knowledge-based AI provides a clearer chain of reasoning behind vehicle behavior.

Decisions can be traced directly back to the rules and logic that produced them, making systems easier to interrogate, validate, and audit.

In practice, this creates several advantages. Engineers gain greater visibility into how systems behave in complex scenarios, helping them identify failure points and improve performance.

It also makes systems easier to adapt for different markets, as local driving rules and compliance requirements can be updated through the reasoning layer rather than retraining or redesigning the entire AI system. This allows manufacturers to scale AV platforms more efficiently across jurisdictions.

From autonomous driving to auditable autonomy

Rather than replacing machine learning, knowledge-based AI acts as a supervisory reasoning layer, applying structured rules and safety logic to monitor and validate vehicle behavior in real time. The result is not simply a vehicle that can act autonomously, but one that can justify its actions.

And the implications extend well beyond autonomous driving. As AI systems are deployed in domains where decisions carry legal, financial or safety consequences, the question of how those decisions are produced becomes as important as the outcome itself.

This is already becoming a defining issue in sectors such as financial services and healthcare, where regulators increasingly expect companies to explain how AI-driven decisions are made.

Ultimately, knowledge-based AI enables AI systems to incorporate defined rules and reasoning into their decision making, rather than relying solely on statistical prediction. In autonomous vehicles, this could take the form of validating maneuvers against traffic laws before execution, but the same principle applies wherever decisions must be explainable, defensible, and auditable.

As AI becomes more deeply embedded in critical infrastructure and public services, the ability to evidence how decisions are made will move from a desirable feature to a baseline requirement across industries.

Proof over performance

The AV industry is often framed as a race to build vehicles that can drive themselves. Increasingly, however, the real challenge is building systems that can explain and justify their decisions in a way regulators, manufacturers and the public can trust.

Knowledge-based AI offers a definitive route to solving that problem. By combining machine learning with structured reasoning, it enables manufacturers not only to improve autonomous behavior, but to explain why systems acted as they did.

For the UK, long-term leadership in autonomous mobility will not be determined by perception systems alone. It will depend on which companies can deliver AI that is demonstrably safe, compliant, and auditable at scale.

We've featured the best AI tool.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

The reported cyber attack involving Canvas and the subsequent ransomware payment will inevitably trigger familiar debates around paying ransomwares.

Most organizations facing ransomware attacks avoid publicly confirming whether a payment was made. Even where payments occur, communications are typically cautious, limited, or deliberately ambiguous.

Admitting to a ransomware payment creates legal, regulatory, reputational, and ethical complications. It can invite scrutiny from customers, insurers, regulators, and shareholders. It may also create concern that the organization has become vulnerable to future extortion attempts.

On one hand, transparency can be viewed positively. Stakeholders increasingly expect honesty during cyber incidents, particularly where personal data is involved. Attempting to conceal the reality of an attack can create longer-term trust issues if details later emerge through other channels.

For many organizations, the decision to pay a ransom is ultimately driven by operational and financial calculations rather than principle alone. If they don’t have things like ransomware protection, backups, or logs it makes it an almost impossible task to recover.

Cyber insurers, legal advisers, and incident response firms may conclude that prolonged recovery, forensic investigation, service restoration, regulatory management, and reputational damage could cost substantially more than the ransom demand itself.

Pressure to restore services

In sectors like education, where downtime directly affects students, exams, coursework, and institutional continuity, the pressure to restore services quickly can become commercially and socially overwhelming.

That does not make payment risk-free or strategically desirable, but it does explain why some organizations determine that the immediate cost of disruption outweighs the uncertainty and expense of a prolonged recovery process.

However, transparency also exposes a more uncomfortable reality within modern ransomware incidents: it does in fact pay to be a cybercriminal.

Yet focusing solely on the ransom payment itself misses the larger issue.

This incident appears to reinforce a wider trend emerging across modern digital platforms: attackers are increasingly exploiting trust itself.

Reports suggest threat actors abused Canvas “Free-For-Teacher” accounts, leveraging a legitimate platform capability designed to support accessibility and adoption. Rather than forcing entry through traditional technical weaknesses, the attackers operated within accepted trust boundaries.

For education providers, this creates a particularly difficult balance. Platforms are intentionally designed to reduce friction for teachers, students, and external collaborators. Accessibility is part of the value proposition. However, the same openness that enables rapid adoption can also create opportunities for malicious actors to blend into normal platform activity.

This is not simply a security engineering issue. It is a governance issue around how digital trust is granted and monitored at scale.

Identity has become the primary security boundary

Cybersecurity strategies historically concentrated on protecting networks, endpoints, and data centers. Increasingly, those controls sit behind identity systems that determine who is trusted, what access they receive, and how quickly they can move through interconnected platforms.

Modern ransomware groups and financially motivated actors increasingly prefer credential abuse, social engineering, and exploitation of trusted workflows because they are often less visible than conventional intrusion methods. A valid account can bypass many of the controls designed to detect malicious behavior.

The challenge becomes even more pronounced in education as, unlike tightly controlled corporate environments, educational ecosystems are inherently decentralized. Institutions regularly support temporary users, external educators, contractors, collaborative learning environments, and remote access requirements. The result is a digital environment where trust relationships are broad by design.

That creates a difficult strategic question for providers and customers alike: how do you preserve accessibility without creating exploitable trust pathways?

The human consequences are often underestimated

Cyber incidents are still frequently measured through technical metrics: records exposed, systems encrypted, or hours of downtime incurred. Those measures rarely capture the wider societal impact.

In education environments, disruption affects students during formative periods of their lives. Exam preparation, coursework submission, academic continuity, and communication channels can all be interrupted simultaneously. Parents and educators face uncertainty around outcomes they cannot directly control.

There is also a more uncomfortable consideration in that educational platforms frequently contain data relating to minors. Even where sensitive information is not immediately weaponized, long-term exposure risks remain difficult to quantify. Personal information tied to younger individuals may retain value for years through identity fraud, social engineering, or future credential abuse.

The emotional dimension of cyber attacks is still poorly understood within many boardrooms because it does not fit neatly into conventional risk reporting.

The due diligence dilemma

Most schools, colleges, and mid-sized organizations cannot realistically perform deep technical assurance assessments against large SaaS vendors. Procurement teams are often left reviewing compliance certifications, security statements, audit summaries, and contractual language that may provide only partial visibility into actual operational practices.

This creates an accountability imbalance.

Customers remain responsible for protecting their own stakeholders and data, yet their ability to validate supplier resilience is constrained by commercial scale and information asymmetry.

That challenge is not unique to Canvas. It reflects a broader maturity gap across the SaaS market.

Many providers publish extensive security documentation, but external assurance still struggles to address practical questions such as: What assumptions are made about “legitimate” users? What controls exist around free-tier or trial account creation?

For customers, obtaining meaningful answers to these questions can be difficult without substantial procurement influence and the result is a market where trust is often inferred rather than verified.

The larger issue beneath the incident

The reported Canvas ransomware payment will understandably drive debate around criminal incentives and incident response decisions. Yet the more strategic question sits elsewhere.

The challenge for organizations is no longer confined to protecting infrastructure from external intrusion. It is understanding where trust is granted, how legitimacy is established, and what happens when a trusted platform becomes the weakest link in a much larger interconnected ecosystem.

That is not merely a cyber security concern.

It is becoming a fundamental business risk question about dependency, governance, and the fragility of digital trust at scale.

We've featured the best cloud antivirus.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

AI is undoubtedly accelerating data scientists' work, but it is also quietly eroding how data science skills are built in the first place. As copilots, automated pipelines, and increasingly capable models take on more of the hands-on work, the role of the data scientist is shifting toward solution design and strategic problem-solving.

Although this may be a welcome evolution for those who have long earned their stripes in the field, it introduces a risk many organizations, as a whole, are underestimating—the loss of repetition and practice that makes this expertise stick.

By reducing first-hand experiences and the challenge of problem-solving, AI-driven automation risks weakening the foundational expertise required for true data science mastery and system-level thinking. According to research from Anthropic, developers who delegated tasks entirely to AI showed weaker learning outcomes even when productivity gains were modest.

For years, developing data science skills meant spending time close to the work. This entailed tasks such as cleaning up messy datasets, performing exploratory data analysis, manual feature engineering, interpreting model outputs, and diagnosing why a model is underperforming.

This kind of hands-on work may not always be efficient, but they are effective. Repeating steps, getting stuck, figuring out what went wrong, and iterating builds intuition and creates a deeper understanding. Repetitive, direct interaction with data, tools, and code transforms knowledge into proficiency, then mastery.

But there’s a tension emerging: the very aspects of AI that make practitioners more productive—automation, speed, and reduced manual effort—are also removing many of the repetitive, hands-on workflows that historically helped data scientists build technical depth and lasting expertise.

Warning signs

The impact on data scientists is immediate and somewhat invisible. When probable answers are just one prompt away, there's less incentive to internalize patterns or build the mental models that enable independent critical thinking and judgment.

Over time, practitioners can complete tasks with AI assistance but struggle to diagnose issues, adapt approaches to unfamiliar contexts, or evaluate whether an AI-generated output is actually correct. In a field where edge cases and ambiguity are the norm, that gap matters.

Without the necessary adaptations to recognize and maintain core expertise, organizations will start to see the warning signs appearing subtly in judgment, troubleshooting, and knowing when to question AI outputs.

How organizations shift their tech teams and data scientists towards thinking in systems as opposed to tasks while reinforcing those core technical competencies will make a difference in ensuring those warning signs won’t progress so far as being clear and obvious negative impacts on the organization.

Hands-on engagement reinforces understanding

This is where organizations need to be deliberate. Not every task needs to be fully automated. The goal isn’t necessarily to slow down AI adoption or force a return to purely manual workflows, but to ensure that as work becomes more efficient, learning doesn’t become incidental.

Here are three frameworks that can help leaders be more intentional about where and how skill practice happens, ensuring AI reinforces learning as well as efficiency:

At the organizational level, dedicate learning time to close the loop between assisted work, knowledge retention, and deliberate practice on fundamentals. If skill erosion is not visible in productivity metrics, then leaders should implement proficiency metrics and periodic assessments.

At the team level, peer and manager reviews are critical to create accountability for independent judgment. This entails reviewing not just outputs but also reasoning, and fostering an environment in which team members challenge each other to explain why things work.

1. At the organizational level, dedicate learning time to close the loop between assisted work, knowledge retention, and deliberate practice on fundamentals. If skill erosion is not visible in productivity metrics, then leaders should implement proficiency metrics and periodic assessments.

2. At the team level, peer and manager reviews are critical to create accountability for independent judgment. This entails reviewing not just outputs but also reasoning, and fostering an environment in which team members challenge each other to explain why things work.

3. At the individual level, the key principle is to preserve engagement with the problem and being deliberate about what parts of the work you stay close to and what you delegate to AI. In some cases, it’s valuable for practitioners to have a dedicated space to engage more directly with the underlying work, such as exploring data without automation or validating AI-generated outputs step by step. Anthropic's aforementioned research supports a specific version of this: Using AI to understand, not just produce.

Fostering these moments of deeper, hands-on engagement across organizations reinforces understanding and long-term capability in ways that passive consumption cannot.

Learning through action makes mastery possible

The AI era is redefining what it means to be a data scientist. As faster tools and more automated workflows unlock new possibilities, teams can focus on more complex problems. But expertise doesn’t emerge from speed alone. It is often best built through experience and a knowledge of fundamentals.

As organizations continue to embrace AI, the challenge is preserving the conditions that build real skills. The “old school” practices that once defined data science—hands-on work, repetition, and learning through friction—are the very mechanisms that enable mastery. Ensuring that work becomes easier without making technology expertise harder to achieve will be critical in the AI-driven future.

We've featured the best AI tool.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

The retail industry is about to lose one of its oldest assumptions: that the shopper at checkout is definitely a human.

30 to 45 percent of U.S. consumers already use generative AI for product research and comparison, and that reliance will inevitably become more pronounced at checkout.

Agentic commerce is beginning to find its way into more consumers’ buying journeys as they look for new ways to shop.

If this new way of shopping maintains its pace, agentic shoppers could make up $190 billion to $385 billion in U.S. ecommerce spending by 2030.

AI agents aren’t only an emerging trend, they are becoming a new class of customer in the commerce ecosystem. But retailers' platforms and websites were not built for this kind of machine-led activity.

There are new pressures building on merchants to rethink and redesign their systems to support autonomous agents and avoid misclassifying legitimate traffic as risky when humans become more hands-off in their buying journeys.

Besides the challenge of becoming discoverable by AI agents, retailers need to be able to verify who is making transactions at checkout when the “shopper” is actually a machine.

That requires understanding which agents are authorized, which ones are malicious, and which ones represent real, valuable customers.

AI agents break the traditional trust model online

The status quo of online retail is being disrupted by AI agents, not because they introduce fraud directly, but because they break the signals merchants have relied on to measure trust for years.

Protocols and identity layers look increasingly different as agents operate in ways that can make them look like suspicious automation under today’s fraud rules. As agents make transactions using APIs rather than typical browsing flows, behavioral analytics loses its predictive power.

In many cases, the usual browsing journey that these brands have used to infer trust simply won’t exist. Retailers can’t assume that the agent is acting on behalf of a legitimate human without proof, so the question shifts from: “Is this user real?” to “Is this agent authorized to act for this user right now?”

The data already points to why this matters: By the end of 2025, online orders driven by LLM referrals were up more than 1,000% year over year. Even so, purchases executed by bots still make up fewer than 1% of all orders.

This is more than a shift in volume. The models that have been trained on human behavior patterns and to recognize bots as bad traffic now struggle when the “user” is a bot with no history and no trusted profile. The data gap creates a dual risk, more fraudulent activities slipping through, and more legitimate orders being declined.

The infrastructure behind agent-safe commerce

Retailers need to start treating AI agents as a new kind of digital customer in their trust systems. This requires an architecture that can authenticate which AI platform or agent is initiating a transaction, rather than treating all machine-driven interactions as anonymous bot traffic.

Ecommerce teams need to focus on providing machine-friendly commerce data with details like product pricing, in-store availability, shipping rules, and return policies that are well-structured, so agents can easily interpret them.

More importantly, they need to distinguish between three categories of activity, malicious automation, authorized agent-driven transactions, and blended human-agent behavior. And they need a way to instantly differentiate between automated threats and AI agents buying on behalf of valuable customers.

The hidden risk: blocking the next wave of customers

It’s a common misconception that the biggest threat retailers are facing is fraud, when the greatest risk is rejecting legitimate orders. What we are seeing now is retailers accidentally blocking agent traffic because it closely resembles typical bot traffic, which means they are losing visibility into how they are being recommended and selected and ultimately into transactions themselves.

Retailers need better classification systems that can separate hostile automation from authorized intent. This requires a more agent-ready commerce stack in five key areas:

Audit the stack for agent readiness: review product data, API accessibility, and machine-readable content to identify where trust breaks across the buying journey

Verify the agent behind the transaction: confirm the identity of the platform or service initiating the order (e.g. ChatGPT, Claude, etc.)

Prove the shopper’s permission: ensure the agent is acting with explicit authorization with controls around limits and categories.

Modernize fraud models for machine-led behavior: optimize classification accuracy, so legit agent-assisted orders aren’t treated like fraud.

Extend trust controls beyond checkout: prepare for agent-driven returns, exchanges, order edits, or support requests.

If retailers implement these steps, they are moving in the right direction to redesign the shopping experience and rebuild the infrastructure beneath it, so they can capture and not block agent-driven demand.

Machine-led commerce is on the horizon

For now, retailers and ecommerce merchants have time to adjust their strategies before agentic commerce enters a mature stage. The shift will start in narrower, repeatable purchase categories, but as the adoption grows, a competitive gap will emerge between retailers that prepared well, and those that didn’t.

To gain that advantage, online brands that modernize their identity, authorization, and risk infrastructure now will be in a better position to support machine-led transactions without adding any friction for the customer behind them.

The retailers that get this right will reduce fraud while capturing a new class of customers. Because even if shopping is done by machines, trust will still need to begin and end with the human customer.

We feature the best credit card processing.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Andy Jassy has enjoyed a long-running and successful career in the technology industry, being a part of the Amazon family for many years – including serving as CEO of AWS for nearly 20 years – before taking over from Jeff Bezos in 2021. In business and life, he believes, there are no shortcuts to success.

Getting hands-on

Jassy coined this turn of phrase when describing AWS in 2017, speaking with CNBC. The company was one of the fastest-growing hyperscalers at a time when the public cloud was becoming a critical part of the technology ecosystem.

Speaking to the success of AWS, Jassy said newer competitors on the market may be able to replicate the technology that the organization offers, but that the years of operational experience accrued made it dominant.

In that vein, Jassy suggested that AWS had been through experiences in dealing with customers that all hyperscalers would have to overcome – and there was no way to answer those questions without going through those same tribulations.

Putting in the hard yards

Compression algorithms are incredibly useful, especially in the age of big data and especially data-hungry large language models (LLMs).

Compression algorithms were the revolutionary tech at the heart of the 2013 HBO show Silicon Valley, and since then many companies have sought to ease the burden, including Google with its TurboQuant algorithm.

As Jassy suggested, however, experience in life and business doesn't work in the same way as data. The quote is perhaps more relevant today than it was then, in light of organizations around the world shedding entry-level jobs.

With businesses cutting the talent pipelines that could have led to years-long careers – much like Jassy's several decades at AWS – it may worsen not just the skills crisis but also create large experience gaps in the business world in the decades to come.

Bill Gates is a one of the stand-out technology pioneers, and has experienced his fair share of hype cycles, having led Microsoft for so many years. When new technologies emerge, it's normal to get excited – but sometimes the rate of progress and the expectations of not just consumers but other businesses fail to match the reality of innovation.

Technology hype cycles

The 'Road Ahead' was a huge bestseller and drew a lot of attention when it launched in 1995. Author Bill Gates, who was CEO and chairman of Microsoft at the time, used the book to opine on various philosophical and technological themes. One highlight in particular was a segment on technological hype cycles and the attitude of people toward innovation.

The first part of the quote – highlighting our fixation on where innovation might lead immediately – alludes to the immediate spike in media attention, excitement and, sometimes, overcommitment. We've seen that time and again over history – with AI a great example of a technology that's undergone numerous hype cycles since the 1960s.

We also fail to capture the reality of compounding gains over many years – with technologies needing not just breakthroughs but ecosystems and support layers. Over the period of a decade or so, we may not realise the small gains made along the way – but zoom and suddenly you realise how different things really are.

Excitement about the future

Exciting technologies today, including AI, threaten to follow in the footsteps of breakthroughs of yesterday, like the internet, smartphones, or even social media.

With expectations heightened, patience narrowing and technology exponentially improving, there's an argument to make that the situation is getting worse, not better, compared with when Bill Gates wrote that sentence in 1995.

Quantum computing is a prime example of how many, including investors, overestimate where the technology might be within a couple of years, projecting the lack of visible improvement forward – to the extent where many write off the tech altogether. Now, however, scientists expect us to build a superpowerful machine by 2030.

Artificial intelligence has transformed enterprise cybersecurity into a machine-speed quickdraw contest.

Today, threat actors routinely use AI and automation to launch sophisticated, multi-stage campaigns that exploit gaps between disconnected security tools.

Once inside a network, modern attacks move laterally across cloud environments, endpoints, and applications within minutes.

Because defensive windows have shrunk from hours to seconds, security teams must rely on AI-driven analytics to correlate threat telemetry and trigger automated remediation before a breach spreads.

To achieve this coordination, many organizations are aggressively pursuing platform consolidation. The logic is simple: by replacing a fragmented patchwork of niche security vendors with a single, unified security platform, a Security Operations Centre (SOC) can centralize its data, simplify management, and orchestrate automated responses more fluidly.

The hidden risks of the single ecosystem

While consolidation can simplify things, it also changes an organization's risk profile. When multiple layers of cybersecurity are interconnected through a single vendor’s control plane, dependencies build up. This level of architectural reliance introduces severe systemic vulnerability.

If your monitoring tools, identity systems, and automated response mechanisms all live under one roof, a single point of failure can paralyze your entire enterprise. A major software flaw, a configuration error, a vendor cloud outage, or a supply chain compromise can trigger a cascading failure that knocks out multiple layers of defense simultaneously.

Furthermore, extensive centralization strips an organisation of its long-term architectural flexibility. Once integrated into a single ecosystem, switching providers or adapting to shifting regulatory and digital sovereignty requirements becomes a massive, cost-prohibitive operational hurdle.

The balanced solution: Hybrid AI architecture

Faced with these challenges, forward-thinking cybersecurity leaders are looking at a happy medium between inefficient platform fragmentation, and total consolidation by adopting a balanced, hybrid approach.

This strategy centralizes AI-driven analytics and detection where shared visibility adds the highest value, while deliberately maintaining strict independence in critical operational zones. A resilient hybrid architecture divides the security environment into two distinct operational mandates:

1. Centralized visibility and detection: Security teams should continue to feed telemetry from endpoints, networks, and cloud infrastructure into a centralized, AI-driven engine such as an advanced SIEM or XDR platform. This allows AI to analyze vast pools of data in real time, map attacker behaviors, and coordinate high-speed incident responses across the enterprise.

2. Isolated control layers: To prevent a total system collapse during a crisis, critical defense layers must remain insulated from the primary detection platform. Two pillars require absolute autonomy:

Identity and Access Management (IAM): Systems controlling user authentication and policy enforcement (like Okta or Active Directory) should not be deeply intertwined with the automated response platform. If an attacker compromises the automated security system, an isolated identity layer prevents them from gaining total, unhindered access to the entire enterprise kingdom.

Backup and Recovery Infrastructure: Disaster recovery tools lose their effectiveness if they rely on the exact same network infrastructure they are designed to restore. Maintaining independent, immutable, and air-gapped recovery layers ensures that even if a ransomware campaign or platform outage takes down the primary network, the business can safely restore operations from a position of absolute control.

Designing for survival

The reality of modern enterprise IT is inherently hybrid, spanning legacy systems, multi-cloud environments, and distributed global workforces. Attempting to force this sprawling complexity into a single security platform is impractical and not without risk.

As artificial intelligence continues to accelerate the threat landscape, the pressure to automate and simplify will only grow. Unified AI ecosystems are essential for operational speed, but true resilience requires architectural balance. Future security strategies will not be judged solely on how quickly they detect a threat, but on how effectively the business can maintain continuity during a catastrophic disruption.

By blending centralized AI intelligence with strategically separated control layers, enterprises achieve the ultimate defensive posture: machine-speed responsiveness without the risk of systemic collapse.

Our rankings of the best cloud backup platforms.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Artificial Intelligence (AI) has become a defining factor in the Managed Service Provider (MSP) market - not only as the service clients want most, but as an operational enabler that MSPs can harness to scale teams, protect margins and deliver better service.

Amongst a backdrop of fiercer competition, smaller deals and a widening talent gap, AI-driven efficiency gains will become essential to ensuring sustained growth.

Recent statistics show that the MSP environment is increasingly defined by tighter deal sizes, cautious buyers and rising delivery cost. Growth hasn’t disappeared altogether, but it is harder to earn.

Not only does securing new business in a crowded market require clearer differentiation. Many MSPs also find that they are continuously having to prove their value to clients, facing dual pressure to maintain high performance levels while operating more efficiently at the same time.

Here are the five key trends currently shaping the market – and how MSPs can adapt:

More competition makes winning deals harder

MSPs need consistent client acquisition to drive revenue. Otherwise, growth slows, revenue shrinks and planning becomes unreliable. However, data shows that winning new customers is getting harder.

Most new clients are switching from another provider rather than buying first-time. This means more providers are competing for the same accounts, with buyers expecting clear proof of value before signing on the dotted line.

When clients have more choices and are more selective, standing out requires more than marketing claims. MSPs should expect their prospects to put heightened scrutiny on pricing, service scope and proven results and be prepared to answer difficult questions.

Those who can demonstrate how their services reduce business risks, improve uptime or increase operational efficiency will be in a stronger position to win new business.

Additionally, structured, profitable and scalable service models will be key to converting demand into revenue.

Average deal sizes are declining

Not only are deals currently harder to close, deal sizes are also declining. As organizations reduce IT budgets, many MSPs are finding that clients will only commit to smaller contracts. Large annual contracts are becoming less common, impacting average monthly recurring revenue across the channel. Taken together, these factors are limiting revenue growth and expansion opportunities for MSPs.

Many are finding that they have to build revenue growth in smaller increments rather than huge account wins. However, data also shows that growth in the market is somewhat uneven: While some MSPs operate with little or no margin, others are experiencing the opposite.

Cost pressure is a recurring theme. Rising labor, tool and IT infrastructure costs are directly constraining growth. Those still posting strong margins have found ways of absorbing higher costs and are managing to weather the storm, with performance in the top margin tiers remaining steady.

The skills gap is having an impact

Talent constraints create additional operational challenges as MSPs are facing difficulties hiring technicians or increasing their workforce to support expanding service portfolios. Often, routine tasks continue to consume large portions of technician capacity.

MSPs should harness automation and AI to reclaim this valuable time. Many are already applying these technologies in high-volume workloads including monitoring, ticketing and alert management, but only a minority so far has achieved broad automation. Moreover, only few have extended AI use to outward growth initiatives such as sales, marketing and client onboarding.

Expanding automation across more areas will reduce manual effort, improve consistency and productivity and allow MSPs to scale operations without increasing staffing at the same pace.

AI is becoming a key differentiator

Looking at where future growth is coming from, AI and automation are emerging as the biggest opportunities. Both have become top client needs ahead of security and backups. Many customers now prioritize AI-driven capabilities over more traditional services.

This creates new opportunities as well as new pressures. MSPs that can turn AI and automation into clear, outcome-focused services will be better placed to stand out in competitive bids and meet evolving client expectations.

However, with providers still working on defining, packaging and pricing their AI related services, it will be some time before AI and automation become meaningful revenue streams. Those that move early to formalize offerings – for example, by packaging AI as a defined service (AIaaS) and showing clear, measurable outcomes from automated workflows – will have a head start in capturing client demand and securing a larger share of the market.

Security remains a reliable growth engine

Cybersecurity and backup services continue to deliver reliable core revenue for MSPs. Clients rely heavily on MSPs for these areas, and for many providers, security represents a key source of income, second only to endpoint and network management.

This is not just driven by growing general security awareness, but also by the ongoing threat and attack levels. With MSPs providing the primary source of cybersecurity advice for their clients, demand is likely to remain high.

Cybersecurity also offers important opportunities for revenue expansion, followed by business continuity and disaster recovery (BCDR). As clients seek to boost their protection against cyber threats, they also invest more in data recovery to improve their resilience against cyber incidents.

To meet client expectations and make the most of the opportunities available, MSPs should treat and market security as a core offering. Providers that strengthen their security capabilities and closely integrate them into their service packages will find it easier to retain and win new business.

This means going beyond basic protection tools such as anti-malware, firewall and network security. Advanced threat protection is essential, while endpoint detection and response (EDR) should be a foundational part of every security stack.

Consolidating security, monitoring and backup tools where possible can remove unnecessary overhead, simplify operations and allow technicians more time to focus on other tasks. AI has an important role to play in this too, delivering smart security insights and automating actions.

As the market evolves further, MSPs will have to work smarter to convert demand into sustained revenue and growth. The providers that will lead in the next cycle are not simply those that add more services. They will be the ones that simplify their stacks, automate intelligently and package emerging capabilities like AI into measurable outcomes.

We feature the best ITSM tools.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Businesses are facing a new challenge - shadow AI. For decades, enterprise teams struggled with ‘shadow IT’, in which employees would bypass procurement processes and approvals to adopt their own cloud platforms and SaaS apps.

Today, employees are rapidly adopting generative AI, AI copilots, and automation platforms outside of controls put in place by centralized IT department. Teams are unleashing the potential AI, but IT isn’t always aware and that’s creating a real risk related to ‘shadow AI’.

The speed of AI adoption is outpacing governance. Yes, employees should absolutely be experimenting with AI. It can automate manual tasks, help employees focus on higher-value work, and drive better decision-making.

The challenge is that companies aren’t always aware of AI usage. Unlike traditional software, AI models and automation tools don’t require significant infrastructure or procurement. Users can adopt new tools right away, without IT’s involvement.

Operational visibility

As adoption grows, AI is getting baked into departmental processes without proper governance or oversight. Companies are no longer just struggling with procurement. They’re struggling with operational visibility.

They don’t know what AI tools their employees are using. They don’t know what data is being uploaded. They don’t know where sensitive data is stored and that leaves companies exposed to operational, compliance, and reputational risk.

Vendor sprawl is creating additional complexity, as it is one of the biggest challenges with unmanaged AI usage. Many companies have woken up to the fact that AI is already being used throughout the business. The problem? They’re trying to simplify operations by adding more AI tools on top of already fragmented technology landscapes.

Every department is procuring different AI platforms. Employees are creating custom automations. Suppliers are dropping AI capabilities into their products with little oversight.

It’s created a disjointed ecosystem of tools, suppliers, and automated workflows. Vendor sprawl is only exacerbated when you throw multiple providers into the mix, such as outsourcing vendors, public cloud platforms, and SaaS vendors, where accountability is divided among parties.

AI blind spots

Add AI into the mix and you’re creating additional blind spots. When an automated workflow breaks, produces incorrect results, or violates compliance standards, who is responsible? The AI model supplier? The underlying software vendor? The automation platform? The person who deployed it? The data source?

If companies don’t have strong governance around AI activity and service integration, these questions will be difficult to answer. AI transformation is an operating model problem, not just tech and many are looking at AI transformation the wrong way. Rather than just trying to deploy AI tech, leaders need to consider how AI tools are used across the business.

Building AI resilience isn’t about using the most AI tools. It’s about building governance, responsibility, and operational resilience into AI activity from the start.

It requires a shift in mentality. IT teams can’t just be gatekeepers anymore. More teams will use AI tools with or without IT approval. Attempting to restrict AI usage will lead to more shadow IT.

IT and IT service management teams need to evolve to focus on service integration, governance, and operational oversight. This includes:

• Creating transparency into AI usage
• Setting responsible AI usage guidelines
• Managing supplier risk
• Integrating AI into operational workflows
• Setting accountability for AI-driven decisions
• Enabling innovation with proper governance

Organizations need end-to-end visibility into AI usage across teams, suppliers, automation tools, and third-party AI services. Without it, there will be cracks that appear in their operational resilience.

Governance will become even more important as regulatory pressure mounts. Governance is only going to become more important as legislators turn their attention to AI. With new regulations like the EU AI Act coming into play, as well as new interpretations of existing data protection legislation, companies are going to be expected to account for how AI tools are monitored, governed, and used.

But most companies are adopting AI long before they have considered AI governance. By the time governing bodies step in to regulate how businesses use AI, businesses will likely be far behind on governance considerations.

Employees may inadvertently share sensitive information through public AI systems. Companies may start using AI-generated content for customer-facing operations without fact-checking or validating quality. Internal decisions may be made by automated workflows with no visibility or auditability.

4 Steps to avoid a shadow AI crisis

It’s not too late for businesses to avoid a shadow AI crisis. However, they need to take action to responsibly manage AI tools and usage.

1. Understand how AI is being used Gain an understanding of what AI tools are being used across the organization, by who, and for what purpose. This includes informal or department-led initiatives happening outside of IT.
2. Define responsible usage guidelines Set clear guidelines for responsible AI use, data practices, supplier risk management and accountability. You don’t need to create restrictive approval processes. But you should create practical guardrails for teams to follow.
3. Treat AI as an operational service AI is increasingly being integrated into business-critical workflows. As such, it should be treated like any other critical service. Define who’s responsible for AI activity, how suppliers are managed, and how security and compliance is enforced.
4. Approach AI governance as a company-wide initiative AI governance shouldn’t be the sole responsibility of IT. Procurement, security, HR, operations, legal, and executive leadership all need to work together.

The reality is organizations don’t need to fear AI. But a starting point is recognizing how unknowingly they’re already losing visibility and control around the technology that’s increasingly powering their business.

We've featured the best AI chatbot for business.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

AI doesn’t make integration intelligent by design. It just makes the gaps harder to ignore.

In mergers and acquisitions, technology doesn’t rescue a poorly prepared integration, it exposes whether two companies were ever ready to operate as one.

Fragmented systems, inconsistent data, weak governance and misaligned access controls: none of that disappears after the deal closes. It sits there, undermining value.

McKinsey’s 2025 State of AI survey found that nearly nine in ten companies now use AI in at least one business function.

Separately, Bain’s 2026 M&A report found that AI adoption in M&A more than doubled last year, with one in three dealmakers now systematically deploying it inside the deal process and across the post-deal operating model.

That acceleration is significant, because many companies are deploying AI before they have resolved whether their data, permissions and governance can support it. In integration work, this becomes visible very quickly.

AI turns M&A fragmentation into business risk

Every acquisition entails some operational overlap that is hard to avoid. The problem is that unmanaged AI use can turn overlapping into an operational contradiction.

Diverse data definitions across the now-integrated businesses can produce inconsistent outputs; different access controls create permission risk; and conflicting governance models leave accountability unclear.

Accounting for duplicate systems that create cost and process drag, AI accelerates these problems rather than resolves them.

When AI draws on inconsistent data across a combined business, its outputs are not obviously unreliable. They look authoritative but misinform decision-making before anyone identifies the contradiction underneath.

Boston Consulting Group analysis found that six in ten companies have yet to show measurable results from AI investments, with poor data quality, inadequate architecture and fragmented governance among the most cited barriers. In M&A, those weaknesses are not inherited once they are inherited twice. Each company brings its own version of the problem, and the merged organisation multiplies every gap.

The risk is not that AI fails outright - it is that AI scales operational fragmentation faster than the business can control it.

The hidden integration problem is governance

Consider two companies that are individually well governed: their permission structures still conflict when merged, data definitions diverge and ownership blurs. Every AI workload layered onto the combined organisation deepens the friction.

This is not a problem of poor management on either side it is structural, and it surfaces the moment companies attempt to operate as one.

After a deal closes, the pressure is immediate. Leadership teams want to combine workforces, standardize systems and start using AI across the new business. Speed is paramount. But AI introduces questions that cannot be deferred.

Who can access which data? Which data is AI allowed to use? Who owns AI outputs? Who audits the decisions AI informs? Which policies govern the new operating environment and who intervenes when outputs are wrong?

These are not questions that resolve themselves over time. Left unanswered, they become embedded in how the combined business operates.

Why this becomes a deal-value problem

This is where deal theory starts to weaken. Synergies depend on shared processes, data and operating discipline. If AI is asked to operate across fragmented foundations, costs become less predictable, integration timelines stretch and time-to-market slows. Security exposure widens as uncontrolled data flows multiply across two estates.

We see this most clearly when companies try to scale AI across a combined business before agreeing on the basic operating rules beneath it. A deal can look attractive on paper, but if the merged organisation cannot produce reliable data flows, consistent governance and stable access controls, AI initiatives will struggle to deliver the value the deal was built on. The gap between what leadership expects and what operations can deliver grows each quarter.

For buy-and-build strategies, the risk compounds with every acquisition. If each new business brings its own systems, data rules and access logic, AI becomes harder to govern with every deal. Without a disciplined approach to operational readiness, the cost of integration escalates faster than the value it was supposed to generate.

What operational maturity looks like in AI-led M&A

The task is not to slow AI adoption. It is to decide what must be standardized before AI is scaled.

For leadership teams, these questions matter most:

1. Can we trust the data? Have the systems and data estates across both companies been fully mapped before any AI workload touches the combined environment?

Without this, AI draws on sources that may conflict, producing outputs that appear reliable but are built on inconsistencies that cannot be traced or corrected.

2. Is ownership clear? Who governs AI outputs, who audits decisions and who is accountable when something goes wrong?

In the absence of defined ownership, errors compound silently and post-incident remediation becomes exponentially more costly than prevention.

3. Is access controlled? Are permissions standardized so that AI draws only on data it is authorized to use, across an environment where the rules are consistent?

Inconsistent access controls are not just a governance risk they create direct security exposure as AI workloads traverse data boundaries that were never designed to be shared.

When these three questions are resolved, cost becomes predictable and the business can scale with confidence. When they are not, every new AI initiative adds risk. The companies that create value fastest from M&A will not be those that apply AI most aggressively.

That means mapping before scaling, standardizing before deploying and resolving ownership before delegating decisions to automated systems. Deal value depends not only on what a business acquires, but on how quickly the combined company can operate intelligently.

AI will not hide operational fragmentation it will put a spotlight on it.

We review and rank the best cloud storage software.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit