With companies of all sizes and across all industries increasingly embracing AI, the technology is rapidly moving from its experimentation phase to showing tangible results.

However achieving such results will rely on having the right partner - especially when it comes to cloud services, where Google Cloud is locked in a titanic battle with its hyperscaler rivals to show its supremacy.

I spoke to Maureen Costello, Vice President UK, Ireland, Sub-Saharan Africa, Google Cloud, to find out more about what the company’s AI tools can do for its customers - and how the UK in particular is leading the way in innovation.

The agentic enterprise is here

“If we were to think about 12 months ago, we were talking about experimentation and PoCs going on, and were they starting to scale - the conversation now will be all about agents and the agentic enterprise,” Costello tells me shortly before going on stage at Google Cloud Summit London 2026 to deliver the opening keynote.

“We're really seeing that shift from using AI to carry out single tasks to actually seeing complex, multi-step processes carried out autonomously by agents - and that's a real shift.”

The company showed off this shift with a number of expanded and new partnerships at its event, including an upgraded relationship with HSBC, which is adopting Google Cloud AI tools and platforms for over 200 use cases across its global services, and the UK government, which will work with Google Deepmind on a new AI-powered tool which should help reduce the time it takes for councils to process householder planning applications.

“Across industries, we're really seeing the ROI starting to come...when you look at where we're working together (with customers) it is in core processes,” Costello notes.

“From a personal point of view, I get excited by seeing the position of the UK globally and within Europe in terms of AI…we’re not just building AI - we are really building and leading the way.”

Costello mentions the “amazing” levels of research going on in the UK, including at Google’s new Platform 37 office in Kings Cross, which features “The Model Garden” a learning space for the latest AI technologies which the company describes as, “a physical marketplace for our best ideas”.

She notes that Google is also working with the UK government to offer upskilling to around 100,000 public servants on AI topics, and is also providing guidance to SMBs looking to take their first steps on their AI journey.

“AI has the opportunity to give a 20% productivity boost to small businesses,” she says, “if you think about that, that’s giving you a day back in a week, to actually do the tasks you need to do.”

Costello mentions Google’s view that AI will look to enhance jobs in terms of how we all use the technology, and highlights how the company is, “working as closely as we can with all sectors of enterprise” to ensure responsible usage.

“The more of those kinds of applications for AI we can show - there's financial ROI, but seeing those benefits land faster and quicker also helps,” she notes, “the more we look at how we can change those types of jobs, the way we can show tangible benefits...that to me is when you can really see the benefits”.

• Google Cloud praises UK market as striding forward with AI
• Age of the "agentic enterprise" is here now, company says
• Google Cloud announces maor partnerships with HSBC, UK government

Google Cloud has laid out its vision for the future of UK businesses enabled by AI, noting that the age of the "Agentic Enterprise" is set to nring in transformation across the board.

“London is rapidly rising and establishing itself as a significant, high-growth hub for global technological innovation,” declared Maureen Costello, Vice President UK, Ireland, Sub-Saharan Africa, Google Cloud at the company's Google Cloud Summit London 206, “this is a market that doesn't just adopt the future, we're going to build it.”

“At Google, we're incredibly proud to be at the beating heart of this British tech revolution - we are not guests in the AI ecosystem, we are deeply rooted in it.”

A "fundamental shift" for businesses

Costello hailed Google Cloud’s recent successes in the UK, noting how, "this momentum is propelling us into a fundamental shift.” This shift, she noted, comes as companies move from the AI experimentation and proof of concepts towards actually realizing true potential.

“The agentic enterprise is happening right here, right now,” she declared, highlighting how every business must now focus on three key pillars when it comes to using AI effectively - culture, responsibility, and sustainability.

This momentum includes a number of high-level partnerships and expansion across top UK firms, including HSBC, which is adopting Google Cloud AI tools and platforms for over 200 use cases across its global services.

This will cover areas such as smarter, AI-driven insights for HSBC wealth management customers, providing more proactive, tailored financial support and real-time advice to customers at every stage of the client journey in a secure way, and using generative AI and agentic AI to build a financial crime architecture that detects risk at an earlier stage, helping to prevent harm and creating a more seamless experience for customers.

“AI is becoming one of the defining technologies of our time, allowing us to create a personalised experience for each customer, delivered in real time and at scale, while keeping human judgement, decision-making, and accountability at the core,” noted Georges Elhedery, Group CEO, HSBC

“A partnership like this one with Google Cloud helps us empower our colleagues with the tools they need to be future-ready, and supports our work in building a simple, agile, faster, and more personal HSBC.”

Google DeepMind will also be widening its work with the UK government, including a new AI-powered tool which should help reduce the time it takes for councils to process householder planning applications down from eight, to four weeks in an average case.

Another tool, Extract, which was announced by the Prime Minister in 2025, is now available to all councils in England, using AI to help planning officers convert decades-old planning documents and maps, sometimes with handwritten notes, into readily useable data in minutes.

“Our planning system remains heavily reliant on cumbersome paper-based processes that consume the time of expert planning officers and cause delays on even the most routine types of application," noted Housing and Planning Minister Matthew Pennycook.

"We are dragging the system into the twenty-first century by harnessing the power of AI to streamline the planning application process, freeing up planners to make quicker and better decisions and reducing unnecessary delays."

• Google Cloud suspended Railway’s account following suspicious activity
• The hours-long outage affected all Railway workloads across all clouds
• Railway is taking ownership of the oversight due to a technical dependency

Platform-as-a-Service (PaaS) company Railway has accused Google Cloud of abruptly suspending its account without prior warning, leading to an hours-long outage.

The company, which counts over three million users who host around 10 million services, APIs and databases, identified an issue at around 10:20pm UTC on May 19 2026, which was only fully rectified around eight hours later at 6:14am.

While the company determined the cause and sought to reinstate services, customers were subject to errors like ‘no healthy upstream’, ‘unconditional drop overload’, login failures and an inability to access the dashboard.

Technical dependency ultimately caused Railway’s outage

Because the network control plane API was affected (which is hosted on Google Cloud), all of Railway workloads across all clouds were affected, returning 503 and 404 errors. Existing workloads stayed up and running for around 15 minutes before caches started to expire.

The Register reports Railway spends an eight-figure sum every year with Google Cloud (potentially upwards of $1m every month), even after having moved parts of its infrastructure into colocation services following 2024 and 2025 issues.

Railway claims it took Google Cloud nearly an hour to engage after the incident happened. “We are livid and still trying to get all the details,” Solutions Engineer Angelo Saraceno said. It’s worth noting that the account got unsuspended by 10:29pm, just nine minutes after issues began.

The company has since published a detailed blog post uncovering everything it knows about the incident, patched together with information from Google Cloud.

The report confirms that the outage was part of a broader, automated sweep by Google that impacted other Google Cloud Platform accounts without prior warning. TechRadar Pro has learned that Google Cloud identified a surge in abusive activity, particularly cryptocurrency mining, across a large number of accounts, and that it had previously warned affected users of suspicious activity and potential lock-outs accordingly.

Exclusive dependency is a bad idea

Railway has taken full ownership of its oversight and has announced immediate changes, including removing the network plane API’s sole dependency on GCP.

“If any of the interconnects go out, there is always a path between the clouds,” Support Engineer Chandrika Khanduri and Agent Experience GM Cody De Arkland explained.

“We have invested in resiliency as a result of prior incidents which have assisted us in dealing with the impact,” the company added, hinting at previous issues. Though it promises to have learned from previous mistakes, it was ultimately Railway’s responsibility to eliminate its GCP-only dependency, which could have prevented widespread impact this time around.

“Your customers don't care whether the failure was Google or Railway; they see your product. Your uptime is our responsibility, and we'll keep delivering on it,” the company concluded.

In an era of greater autonomy after years of experimenting with passive artificial intelligence, Google Cloud Global Director of Healthcare Aashima Gupta says it's no longer a question of adoption – it's a question of implementation.

Speaking in an exclusive interview at Google Cloud Next 2026, Gupta outlined how AI has rapidly embedded itself into clinical workflows – and not by top-down pressure. Instead, frontline demand has been a huge influence in how AI systems get delivered across the industry, with tools designed to reduce worker fatigue at the cause.

In one example, automating nurse handoff notes saves minutes per interaction across tens of thousands of daily transitions, reclaiming significant time for actual patient care.

Google's vision for AI in healthcare is almost entirely administrative

But the real transformation is yet to come, lying in what Gupta describes as healthcare's "agentic moment." Rather than just assisting humans, AI agents are beginning to orchestrate entire workflows, reasoning across systems, adapting to context and executing multi-step processes.

"These are agents – they don't get tired," she added, highlighting the power of autonomy in handing repetitive, administrative tasks while clinicians focus on high-value, human-centered care.

Agentic is the name of the game in an industry still dominated by legacy systems and physical paper records. Gupta argues that modern AI agents can now stitch together systems like electronic medical records, imaging systems, labs and insurance platforms.

But while workers are increasingly accepting of AI assistance having experimented with consumer products in their own time, Google Cloud knows the real battle to widespread adoption is in trust, not capability.

Data residency, audibility and role-based access controls are all key to making AI succeed within healthcare – one of the most regulated industries, but also one of the industries likely to benefit the most from correctly deployed AI.

Gemini doesn't want your patient records, but Google wants to help you connect to them vis your surgery

Looking ahead, Gupta envisions a future where every patient has access to an AI assistant through their own healthcare provider, hospital or clinic. Gemini's power isn't enough, she explained, noting that Google doesn't have access to your personal records (nor does it want to).

But for hyper-personalized healthcare recommendations that could prevent visits to the clinic altogether, purpose-built AI chatbots are very much the goal.

And as for the smaller healthcare providers, Gupta's message is clear: "Be in the arena." Start experimenting now with early pilots around high-frequency, low-risk workflows to gain internal expertise, and build out from there.

Ultimately, Gupta sees AI as a capacity engine for clinicians and not a replacement for workers in an industry defined by human interaction.

"The era of the pilot is over," Google Cloud CEO Thomas Kurian declared at the company's annual Next conference earlier this mont, "The era of the agent is here."

With more than 32,000 attendees on the ground and over 260 announcements made, spanning infrastructure, data, security and applications, the event marked one of the company's biggest moments yet.

The message was clear: agentic AI is here, and Google Cloud wants to offer solutions across the whole stack.

What agentic AI means for Google

Central to this year's Next's narrative was a shift in how Google envisions organizations interacting with AI as they shift from generative to agentic AI.

For the past two years, Google Cloud explained, most of the industry has focused on passive, copilot-style assistants to help employees write, code or analyze faster.

This year, AI is moving beyond assistance to become an active participant in business operations, often totally separated from a human worker. Agents are being positioned as "intellectual peers" – not just tools – capable of reasoning, coordinating and executing tasks.

Speaking in a pre-recorded message at the event, Google CEO Sundar Pichai revealed that around 75% of the company's new code is AI-generated, but we also now know that around 75% of Google Cloud's customers are now using AI in some capacity.

To support growing interest, the company outlined what it's calling its agentic enterprise blueprint, which rethinks the full stack of enterprise IT architecture. Hence the seemingly endless new product announcements and updates.

What agentic AI means for enterprises

When he stated that the era of the pilot is over, Kurian means to say that investing in a unified platform of data, models, infrastructure and applications is key, rather than layering AI on top of existing, legacy systems.

Speaking to a group of journalists, Agentic Data Cloud VP and GM Andi Gutmans explained that "humans can only click so many times a day." With agents effectively removing that ceiling, operating around the clock and at scale, enterprises must totally rethink their data foundations to maximize AI ROI.

Gutmans stressed that, even today, around 90% of all enterprise data is unstructured 'dark' data that's historically been underused.

Agentic technologies promise automated date discovery, interpretation and enrichment, and Google Cloud is not only at the forefront of this, but it's making strides to meet customers where they are.

The company explicitly acknowledged that enterprises run multi-cloud and hybrid environments beyond just Google Cloud, and it has a clear vision to activate all enterprise data wherever it is, not to "move everything to Google Cloud."

At the same time, company leaders acknowledged that this transition introduces new complexities such as agent management, governance, security and cost, all while systems become highly interconnected and interdependent. Something it aims to tackle with the new Gemini Enterprise platform, which we've covered separately.

At the end of the day, agentic may be here but companies may not be agentic just yet, and Kurian's message serves as a warning message for them to reconsider their stack from the ground up rather than continuing to build and add layers of complexity on top of it.

• The new Gemini Enterprise Agent Platform is an end-to-end building and deployment tool
• Google's clearly committed to interoperability with third-party model and tool support
• Even non-technical workers should be able to build their own AI agents

Google Cloud has unveiled Gemini Enterprise, which has evolved into a single interface where users can interact with their AI agents just as they would their Workspace apps.

Core to the announcement is the brand-new Gemini Enterprise Agent Platform, described as an end-to-end development platform for building, deploying and managing agents at scale.

Designed to be as simple to interact with as the rest of the Google Workspace suite, interoperability was also a core message at Google Cloud Next 2026 and is pivotal to how the new Agent Platform works.

Google wants Gemini interactions and management to be as easy as possible

Google described the platform as a model-agnostic ecosystem that lets users either access Google's own models or third-party alternatives for maximum flexibility.

Agents can also share context across systems, apps and workflows to help make them more effective, with the Gemini Enterprise platform acting as a central monitoring and auditing tool.

Gemini Enterprise Senior Director of Product Maryam Gholami also noted that companies have shifted from generative AI and agentic pilots to full-scale agentic deployments, noting the requirement for always-on automation.

"Companies are ready to build their agentic task force, but this demands doing so within a secure and governed environment," Gholami shared.

Described by Google Cloud CEO Thomas Kurian as 'the primary environment where your business actually operates, the Gemini Enterprise app has also been reshaped with an improved Agent Designer, so non-technical workers can create workflows with reusable Skills with natural-language prompts.

A screenshot of the updated interface shows a visual, branch-based builder that can handle 'if this, then that' type splits and human-in-the-loop checkpoints where approvals may be required.

By unifying the whole stack agentic AI stack, Google wants to be much more than a tool provider, pushing back against rivals with a fully end-to-end management and deployment platform.

• OpenAI declares Microsoft a "primary cloud partner," but it's no longer an exclusive partner
• The "greater predictability" will help the ChatGPT maker to "scale"
• Recent shifts continue to point toward an IPO

OpenAI has declared an end to Microsoft's exclusive cloud rights, noting it plans to use other cloud providers like AWS and Google Cloud going forward.

The ChatGPT maker noted Microsoft will remain its "primary cloud partner," whereby it will select Azure as its first choice wherever suitable, but the company now promises to "serve all its products to customers across any cloud provider."

As part of the changes, Microsoft will retain access to OpenAI models via a license that lasts until 2032, but it's not unique to Microsoft as OpenAI looks to step away from this exclusivity.

OpenAI and Microsoft are no longer exclusive

"The greater predictability in the amended agreement strengthens our joint ability to build and operate AI platforms at scale while providing both companies the flexibility to pursue new opportunities," OpenAI wrote in its announcement.

As for Microsoft, it's no longer set to pay a revenue share to OpenAI as it has done under the exclusivity deal, but it will remain a major shareholder in the company. OpenAI will continue to pay a revenue share to Microsoft until 2030 at the same percentage, with a cap in place.

It marks around five years of Azure exclusivity, whereby all major products were trained on, and distributed via, Azure. With the contract effectively blocking OpenAI from using AWS, Google Cloud and others, changes started to become apparent in 2025-2026, driven by the Oracle-backed Stargate Project and, a year later, a strategic partnership with Amazon.

Increasing independence could play into OpenAI's strategy to go public, rumors of which have been circulating for a number of months.

• Google unveils next-generation TPUs – splits off into two series, 8t and 8i
• 8t superpods can deliver 121 ExaFlops, up from 42.5 last year
• 8i delivers 3x more SRAM and increased HBM

Google Cloud has announced its eighth-generation Tensor Processing Units (TPUs) designed specifically for the agentic shift we're seeing within AI at the moment.

Revealed at Google Cloud Next 2026, the upgrades focus on longer context windows, multi-step reasoning and responsiveness at scale, thus its cloud infrastructure is being rebuilt to support persistent memory, continuous inference and multi-model workloads.

This year, we're seeing two distinct TPUs designed to support massive HBM scaling, with Google Cloud placing an emphasis on memory bandwidth as much as compute.

TPU 8t and 8i target trillion-parameter training in million-chip clusters

The first of two TPUs, 8t, has been optimized to be distributed across huge clusters for training foundation models. With around an 80% year-over-year improvement in performance per dollar, the company says it will train trillion-parameter models more efficiently.

Google Cloud explained that a single TPU 8t superpod can scale up to 9,600 chips, delivering 2PB of shared HBM and 121 ExaFlops of compute. For comparison, last year Ironwood was rated at up to 9,216 chips in a superpod and 42.5 ExaFlops.

Google Cloud also warned of "the latency wall" we face in an always-on agentic era, hence the launch of 8i, a second chip which serves as a post-training and inference engine.

TPU 8i sees around a 3x increase in on-chip SRAM to 384MB as well as 288GB of HBM, with pod size now up to 1,152 chips from 256, delivering 11.6 ExaFlops of performance (up from 1.2 ExaFlops).

As for energy and thermal efficiency, Google Cloud boasts of up to 2x better performance-per-watt over Ironwood, the predecessor.

"We['ve] innovated across hardware and software to enable our data centers to deliver six times more computing power per unit of electricity than they did just five years ago," SVP and Chief Technologist for AI and Infrastructure Amin Vahdat explained.

General availability for Google Cloud customers is expected in the coming months, and naturally, TPU 8t and TPU 8i will be at the forefront of the latest Gemini models.

The company also sees the eighth-gen hardware playing a role in developing the next frontier models by distributing training beyond a single superpod using Pathways and JAX to unlock scaling beyond one million TPU chips per any single training cluster – something execs confirmed at the event is currently entirely theoretical (but technically possible), with the TPUs yet to be made available at such a scale.

That's a wrap on Google Cloud Next!

It was a packed few days of live coverage straight from Google Cloud's biggest annual event.

The first day kicked off with the main keynote covering tens of separate announcements, and the rest of the conference provided opportunities to explore them in more detail so we can bring you all the latest news from Google Cloud this year.

If you want to check out what we saw at the show, you can catch up on our live blog below.

The venue is filling up with around 32,000 attendees as Google Cloud gets ready hold its biggest event of the year, packed with new product announcements and more.

Good morning from Google Cloud Next. The first keynote kicks off at 9am PT, so join us in a little over an hour for all the details as they roll out.

In true tech conference style, a live DJ set is warming the audience up ahead of today's announcements

"How do you move AI into production cross your entire enterprise?" Google Cloud CEO Thomas Kurian asks as he welcomes the audience.

Spoiler alert, he's already given us the answer – it's a unified stack. That's what we'll be hearing about over the next two days.

Around 75% of all new Google code is AI-generated, Google CEO Sundar Pichai says in a pre-recorded video. Around a year ago, we reported that AI was accounting for around 25% of Google's new code (and a third for Microsoft).

The focus of today's keynote and this year's announcements will be Google Cloud's new Gemini Enterprise Agent Platform, described as an "end-to-end system for the agent era"

We're also hearing about Gemini 3.1 Pro, Gemini 3.1 Flash Image (Nano Banana 2), Veo 3.1 Lite and Lyria 3 Pro

And name drop... Google Cloud is reminding us about the landmark Apple partnership, which will drive this year's much-improved Siri

Name drops are continuing as Thomas Kurian reminds us that Google Cloud partnered with NASA/Artemis II on flight readiness with agentic AI

Projects in Gemini Enterprise, coming soon, serves as a shared workspace for teams and agents

"The era of the pilot is over, the era of the agent is here," Kurian declares as he welcomes colleague Erica Chuong for a demo.

For another live demo, 3x Olympic Gold Medallist and snowboarder Shaun White is showing us how Google Cloud and Gemini can analyze the details of a video for detailed analytics.

Learning a trick on the mountain is one thing, White says, but visualizing and analyzing the physics unlocks something that's never been possible before.

Oh, and of course, the convention center filled with "snow" for White's entry.

Next up, TPU 8t and TPU 8i are being announced

TPU 8t promises 3x performance per pod, support for up to 9,600 TPUs via 3D torus topology, 2 PB of shared bandwidth memory (enough to hold the entire digital collection f the Library of Congress 100x over)

As for TPU 8i, there's now support for up to 1,152 TPUs per pod via Boardfly Topology. We've broken down the "memory wall", Google says

Of course, we've got the fuller details for Google's two new TPUs coming later on

Chief Product and Business Officer Karthik Narain is here for the next launches, kicking off with the Knowledge Catalog, Smart Storage for agent-ready data with instant tagging and zero manual data engineering, and Deep Research Agent

Sticking on the data side, Google is launching the Cross-Cloud Lakehouse that's "completely borderless"

"No more vendor lock-in, just freedom," Narain declares

Just weeks after we learned that Google Cloud completed its acquisition of Wiz, we're now getting an insight into what this acquisition means for Google Cloud.

In a quick demo on stage, we see Gemini Enterprise for Customer Experience handle complex workflows with product queries and even language changes. For the consumer it means that conversational AI can be used via phone support, with the AI agent having access to all the necessary resources behind the scenes.

Yulie Kwon is on stage now to announce the general availability of Workspace Intelligence – deeper context for information gathering with situational awareness and personalization. It turns fragmented information into a clear path forward, she adds.

Additionally, Rapid Enterprise Migration, in Preview, is now up to 5x faster to move an entire organization from M365 thanks to big work on interoperability

Next up, we're heading to a breakout to hear what's new with Gemini, from Google DeepMind

While that's it for the live keynotes, we're still on the ground jumping between breakouts and speaking to Google execs for all the latest details on this year's long list of announcements.

Good morning from the second of two days at Google Cloud Next 2026. This morning, we're having an exclusive session with Google Cloud CEO Thomas Kurian to hear about some of the announcements in greater detail.

In a room of journalists, Kurian is speaking about Google Cloud's European sovereignty principles, model distillation, the increasing pace of product announcements and internal development, and quality control.

Google Cloud is very proud of its $32 billion acquisition of Wiz. Kurian is telling us about how its technology will help to automate cybersecurity – AI can simulate red teams (attack simulation), blue teams (key to identifying the vulnerabilites), and green teams (for figuring out the fixes).

Speaking with TechRadar Pro in an exclusive interview, Google Cloud Global Director of Global Healthcare Solutions Aashima Gupta reiterated a message we've been hearing a lot this year, not just at Google Cloud Next, but also from Nvidia CEO Jensen Huang (speaking at Adobe Summit earlier this week).

Artificial intelligence will handle the chore, she explained, while workers will be able to focus on the actual purpose of their jobs.

For healthcare clinicians, this means they'll be able to spend more time taking care of patients – but we'll explore this in greater detail separately.

Speaking with EMEA journalists in a breakout about defining ROI in the world of agentic AI, Oliver Parker, VP for AI GTM and Alex Rutter, EMEA MD for AI explain that it's difficult to standardize this. Instead, each organization should consider what they want from AI as the process becomes far more personalized.

In about half an hour, we'll be getting a hands-on, real-time demo of some of the new AI tools in Google Workspace to see how Google Cloud has been tackling some real use cases

Now to see some of the latest additions to Workspace in practice!

We're seeing Workspace Intelligence at play, pulling context from multiple places across Drive, Gmail and more to generate new documents, using the same formatting as a reference document.

Gemini can also address comments from co-workers in Workspace apps like Docs – actions are taken and produced in a 'preview' in-line for the user, who can go ahead to accept or reject them.

Another demo, this time in Sheets, shows Gemini being able to fill out cells based off context, with no additional prompts needed

Next up, we're seeing how a prompt can be entered into the Drive search bar to generate an AI Overview

A quick hop and a jump now over to a 'chip chat' session with Mark Lohmeyer (VP/GM, AI and Computing Infrastructure) and Leo Leung (Director of Product Management AI & Computing Infrastructure) to hear about Google's new eighth-gen TPUs this year. We'll be covering these separately soon

Next up, it's a hands-on demo for us to build an agent with Gemini Enterprise

That's more or less a wrap on our time at Google Cloud Next, but keep an eye out on what's to come as we break down some of the product announcements, including the eighth-gen TPUs and new Workspace improvements. Thanks for joining us!

• The global cloud market has been growing by 20% or more for six consecutive quarters
• 2025 cloud spend was very nearly $400 billion, led by three global hyperscalers
• Omdia predicts a further 27% growth in 2026 as demand continues

New figures have claimed global cloud spend in the final three months of 2025 rose 29% year-over-year to $110.9 billion, making it the sixth consecutive quarter not just of any growth, but of 20%-or-higher growth.

Omdia, the company behind the data, explained AI's shift from experimentation to production deployment is to blame for the huge growth, with spend expanding far beyond the already-high demand for CPUs and GPUs to storage and networking too.

With quarterly cloud revenue coming in at $90.9 billion in Q1, $95.3 billion in Q2 and $102.6 billion in Q3, this brings the full-year spend to a total of $399.6 billion – a year-over-year growth rate of 24%.

Global cloud market keeps on growing

But as for the individual companies that make up the market, nothing has changed with three leading hyperscalers remaining unbeatable. AWS continues to account for around one-third (32%) of the market, with Microsoft Azure (22% share) and Google Cloud (12% share) following in second and third place.

Though Google trails far behind its competition, its 50% growth is far bigger than those of Azure (39%) and AWS (24%). Google is the third of three $3 billion companies, behind Apple and Nvidia.

"As AI continues to raise infrastructure requirements while constraints remain, vendors that can expand in a more targeted and efficient way will be best positioned to lead in the next phase of competition," Senior Director Rachel Brindley wrote about the months that lay ahead.

Omdia also noted that AWS ended the quarter with a $244 billion backlog and Google Cloud with a $240 billion backlog (up from $157.7 billion in just three months), indicating sustained demand.

Looking ahead, Omdia is predicting a further 27% growth in cloud infrastructure spending in 2026, bringing total yearly spend to more than $500 billion.

• Attackers abused Google Cloud Application Integration to send phishing emails from legitimate Google domains
• Emails mimicked Google notifications, redirecting victims through trusted services
• Nearly 3,200 businesses targeted; most victims in U.S. manufacturing, tech, and finance sectors

Legitimate Google services are, once again, being abused in phishing attacks, successfully tricking targets into clicking malicious links and giving away their login credentials.

In a newly released report, cybersecurity researchers from Check Point said they’ve seen almost 10,000 emails, being sent to roughly 3,200 businesses in a span of two weeks.

All of the messages were sent from the email account noreply-application-integration@google.com, meaning the attackers were abusing Google Cloud Application Integration.

Targeting manufacturing in the US

This is a managed Google Cloud service that connects applications, APIs, and data sources without needing to write custom code. It lets organizations automate workflows between cloud services, SaaS apps, and internal systems using prebuilt connectors, triggers, and actions. Emails generated through Google Cloud Application Integration often originate from Google-owned infrastructure and domains, meaning they’re sent as part of an automated workflow and can inherit Google’s strong sender reputation.

In phishing campaigns, threat actors can create or compromise a Google Cloud project and configure an integration workflow that sends emails via Gmail APIs or other connected email services. In other words, this is simple abuse - not a breach in Google’s infrastructure.

To make the emails seem even more plausible, the attackers made sure the messages closely followed Google’s notification style, language, and formatting. The most common lures include pending voicemail messages, or notifications about being shared a document.

The link shared in these emails leads to storage.google.cloud.com which is a trusted Google Cloud service. However, it then redirects to googleusercontent.com, where they need to pass a fake CAPTCHA built to block security scanners. Finally, the victims are redirected to a fake Microsoft login page, where they can be tricked into giving away their login credentials.

The majority of the victims were in the United States (48.6%), working in manufacturing/industrial (19.6%), technology/SaaS (18.9%), and finance/banking/insurance (14.8%).

Google told Check Point “several phishing campaigns” abusing Google Cloud Application Integration were already blocked.

“Importantly, this activity stemmed from the abuse of a workflow automation tool, not a compromise of Google’s infrastructure. While we have implemented protections to defend users against this specific attack, we encourage continued caution as malicious actors frequently attempt to spoof trusted brands. We are taking additional steps to prevent further misuse,” Google concluded.

• Google Cloud's Partner Network is being upgraded for all types and sizes of partners
• There's a refreshed focus on outcomes, with some useful simplicity and automation improvements, too
• The introduction of an "intentionally selective" Diamond tier will reward the highest achievers

Google Cloud has introduced a new partner program for the company, set to launch during the first quarter of 2026.

The new and upgraded Google Cloud Partner Network goes from, "measuring program work to valuing genuine customer outcomes," which means sales, service delivery and innovation will all be rewarded.

All partner types and sizes (ISVs, RSIs, GSIs and more) will be included, and Google has committed to adding six months of transition to help partners adapt to the change.

Google Cloud's channel program has been updated and simplified

In a blog post, Google Cloud Channels and Partner Programs VP Colleen Kapase explained the new program has been built around three core pillars – simplicity, outcomes and automation.

Beginning with simplicity, Google is moving away from traditional requirements like business plans and customer stories to focus on partner contributions, like pre-sales influence, co-innovation and post-sales support. Google Cloud also promises to reward investments in skills, real-world experience and successful customer outcomes.

For outcomes, the company is adding a new Diamond tier above the previous Select and Premier models (which live on under the changed format). Diamond is described as "intentionally selective," designed for partners who consistently deliver exceptional outcomes.

Finally, we have the biggest change to Google's Cloud Partner Network. The Hub has been given a makeover for 2026, and will now use AI and automation for tracking things like competency and customer engagements, allowing partners to reduce the amount of manual reporting they have to do.

For existing partners, it means that they'll transition onto the new system over the next six months, and they're not guaranteed to retain their current tier. A new fight will also emerge for partners to reach the top-tier Diamond status, which might grant them access to higher visibility.

• AWS and Google Cloud announces closer ties
• New multicloud interoperability tool looks to lessen headaches across the board
• AWS says more cloud partnerships set to come soon

As many big businesses will confirm, having multiple cloud systems on board can often be a major headache, particularly when it comes to interoperability.

AWS is looking to solve these issues with a new service it claims can remove many of the roadblocks in traditional multicloud networking.

Revealed at its AWS re:Invent 2025 event, the new AWS Interconnect - multicloud will look to allow better integration between AWS and other service providers - beginning with Google Cloud.

AWS and Google Cloud partnership

The two cloud giants have revealed a new open specification for network interoperability using both AWS Interconnect - multicloud and Google Cloud’s own tool Cross-Cloud Interconnect.

The companies say this step will allow customers to establish private, high-bandwidth connectivity between the two providers much faster and easier than before - as previously, customers connecting different cloud workloads had to either use public connectivity with no bandwidth guarantees, or build complex private connectivity systems.

AWS Interconnect - multicloud looks to offer a fully managed, cloud-to-cloud experience using pre-built capacity pools to allow organizations to create connections and adjust their bandwidth as needed.

Service providers are able to fully manage the infrastructure, with built-in resiliency and streamlined support giving customers the ability to remove the overhead of managing physical devices or virtual routing objects from their multicloud networks.

“This collaboration between AWS and Google Cloud represents a fundamental shift in multicloud connectivity," noted Robert Kennedy, VP of Network Services, AWS.

"By defining and publishing a standard that removes the complexity of any physical components for customers, with high availability and security fused into that standard, customers no longer need to worry about any heavy lifting to create their desired connectivity. When they need multicloud connectivity, it's ready to activate in minutes with a simple point and click.”

• Google’s AI VP and company CEO warn of an “intense” year ahead
• The company wants to double AI serving capacity every six months
• Reducing reliance on third parties could help solve some cost and efficiency concerns

Google’s AI and Infrastructure VP, Amin Vahdat, has reportedly warned employees that the company must double serving capacity every six months in order to keep up with demand for AI tools.

CNBC reported the news hit staff in a company all-hands, during which Vahdat revealed that Google would need to scale “the next 1000x in 4-5 years.”

Vahdat noted all of this was needed while maintaining the same cost and power consumption, spelling out a challenging future of both colossal capacity increases and equally mighty efficienty improvements.

Google needs to double AI progress every six months

Vahdat explained the 1000x scale, targeted for around the end of the decade, would need to come at “essentially the same cost and increasingly, the same power, the same energy level.”

Clearly, the roadmap consists of multiple elements. Google continues to work on expanding its infrastructure, like AI and cloud data centers, but it’s also rolling out more of its own hardware (like TPUs) to reduce reliance on third-party companies. Nvidia has profited hugely from that, for instance.

Google’s seventh-generation TPU, Ironwood, claims a 30x power efficiency boost over 2018 models.

And on those third-party concerns, many Nvidia chips are being flagged as ‘sold out,’ per The Verge, which has slowed down some rollouts across the industry, including Google’s own AI features.

Separately, Google CEO Sundar Pichai has also warned 2026 would be “intense” due to AI competition and compute demand. Google has publicly acknowledged AI bubble concerns, but Pichai deems underinvesting in AI even riskier.

• NATO has selected Google Cloud to power JATEC
• Google Distributed Cloud is an air-gapped, sovereign cloud for maximum security
• Google Cloud revenue skyrocketed last quarter, pushing it to be the third-most valuable company globally

NATO’s Communication and Information Agency (NCIA) has signed a multimillion-dollar deal with Google Cloud to enable it access to sovereign cloud capabilities amid the organization’s digital modernization efforts.

Google Cloud and NATO said the “strategic partnership” would help NATO strengthen its data governance and give it access to secure cloud and AI tech.

More specifically, NATO will be using Google Distributed Cloud (GDC), which is an air-gapped sovereign cloud solution designed specifically for highly regulated industries, promising disconnected and secure environments.

NATO strikes a deal with Google Cloud

“This empowers organizations to run modern AI and analytics workloads on their most important data, unlocking valuable insights while maintaining absolute operational control and meeting the strictest digital sovereignty requirements,” the two wrote in the announcement.

GDC allows NATO to have control over data residency to keep within its strict sovereignty and compliance guidelines. Google stressed that GDC keeps “NCIA’s highly sensitive data… under its direct control and within NATO’s sovereign territory.”

“This partnership will enable NATO to decisively accelerate its digital modernization efforts while maintaining the highest levels of security and digital sovereignty,” Google Cloud EMEA President Tara Brady wrote.

Google Distributed Cloud is the chosen provider for NCIA’s Joint Analysis, Training and Education Center (JATEC), a NATO-Ukraine organization that handles data from the Russia-Ukraine war.

“Through this collaboration, we will deliver a secure, resilient and scalable cloud environment for JATEC that meets the highest standards required to protect highly sensitive data,” NCIA CTO Antonio Calderon wrote.

Google and NATO have kept the finer details of the deal under wraps – it’s unclear how many millions the deal could be worth, and for how long it will be in place. Still, it’s good news for Google – a company that continues to gain market share in the cloud sector.

Google Cloud revenues increased 28% last quarter, which is more than double the 12% rise in revenues Google as a whole saw.

In fact, strong fiscal performance saw Google recently overtake Microsoft in terms of market cap. Valued at $3.617 trillion, it trails only behind Apple and Nvidia (both over $4 trillion).

• AWS, Microsoft Azure and Google Cloud occupy more of the cloud market than before
• Analysts claim AWS position is slowly dropping, Microsoft and Google climbing
• Annual growth will continue at 20-25% for five years

New Synergy Research figures have claimed around 60% growth in the global cloud infrastructure services market over two years, with spend reaching $107 billion in the first quarter of 2025.

The data also reveals the growing dominance of the big three providers – AWS, Microsoft Azure and Google Cloud - now hold a combined 63% of the market, up from 62% in 2024 and 61% in 2023.

However, despite being the biggest of the three, Amazon’s position has been declining steadily since the second quarter of 2022, and it could just be time separating it from first spot with both Microsoft and Google on the rise.

AWS could be on track to drop a position in the cloud market

The three companies drew attention to AI’s impacts on cloud growth.

Though still a long way behind AWS (29%) in terms of market share, Microsoft Azure (20%) has shown a generally consistent upward trend. Google (13%), further behind still, has also been growing steadily. It’s now approaching 4x the size of Alibaba, which resides in fourth place.

“Its share is showing gradual erosion as Microsoft and Google continue to close the gap, yet it remains striking how effectively Amazon has maintained its leadership position,” Chief Analyst John Dinsdale explained.

Looking ahead, Synergy anticipates the average annual growth rate in cloud revenues to remain above 20% for the next five years. Last quarter, they stood at around 25%.

In the three months ending September 30 2025, AWS saw a 20% year-over-year increase in revenue to $33 billion. Microsoft trumped Amazon in the same three months, with a 26% rise in revenue to $49.1 billion. Though smaller, at $15.2 billion, Google Cloud saw the biggest rise of the three, at a staggering 34%.

• Developers can now build AI agents that ‘self-heal’ with Google Cloud
• Go has been added to the list of supported programming languages
• Dashboards improve observability and security

Google Cloud has rolled out a series of updates to its Vertex AI Agent Builder in the hope that developers continue to build, test and deploy AI agents on its platform, with an emphasis on faster building.

The company has updated its Agent Development Kit (ADK) so that developers can use prebuilt plugins, such as a new one designed to help agents ‘self-heal’, meaning that they can become even more autonomous.

Go has also been added as a supported programming language, on top of existing Python and Java support, which Google stresses increases accessibility.

Google wants developers to build their AI agents even more quickly

Once agents are build, a single command via the ADK CLI will see them get deployed. “This is a major upgrade to help you move your agent from local development to live testing and production usage quickly and seamlessly,” Product Management Director Mike Clark noted.

After their AI agents have been deployed, developers will now gain extra observability through dashboard enhancements, including oversight over token usage, latency, errors and tool calls.

A dedicated playground will also allow them to test and debug faster, while the traces tab helps visualize agent action sequences to diagnose issues with clear paths.

PayPal Principal Engineer for AI Nitin Sharma praised Google Cloud’s Vertex AI Agent Builder for its support across “inspect[ing] agent interactions, follow[ing] state changes and manag[ing] multi-agent workflows.”

But with quicker development and deployment could come heightened security risks, so Google has addressed these too.

“Many developers are prototyping AI agents, but moving to a scalable, secure, and well-managed production agent is far more complex,” Clark added.

Model Armor will screen tool calls and agent responses, checking for prompt injection, while the Security Command Center serves as an agentic asset inventory.

• Anthropic strikes deal worth “tens of billions of dollars” with Google Cloud
• Claude maker will add more Google TPUs to its Nvidia GPU and Amazon Trainium portfolio
• Amazon remains Anthropic’s “primary training partner and cloud provider”

Anthropic has signed a multi-billion dollar deal with Google which will see the AI firm gain up to one million Google Cloud tensor processing units (TPUs) for training and running its large language models (LLMs).

The deal, worth “tens of billions of dollars” according to Anthropic, will mean the company could have well over a gigawatt of compute capacity by 2026.

The expansion comes after a successful year for Anthropic, which has seen the number of customers worth $100,000+ in run-rate revenue grow nearly 7x in the past year thanks to the ongoing AI boom.

Anthropic boosts capacity

Anthropic CFO Krishna Rao notes the move is a continuation of the partnership between it and Google Cloud, which was likely chosen because of its price performance and energy efficiency – not to mention the existing positive relationship between the two.

Because of the power delivery design of Google’s TPUs, they’re typically more power-efficient than Nvidia’s GPUs, which are synonymous with other AI tech such as Anthropic competitor OpenAI.

Such a large deal is probably a good sign for Google, then, which wants to challenge Nvidia’s market dominance.

“We are continuing to innovate and drive further efficiencies and increased capacity of our TPUs, building on our already mature AI accelerator portfolio, including our seventh generation TPU, Ironwood,” Google Cloud CEO Thomas Kurian explained.

However, Google isn’t Anthropic’s only partner. Nvidia’s GPUs and Amazon’s Trainium also play an important role in the company’s multi-vendor strategy, and indeed Amazon remains Anthropic’s “primary training partner and cloud provider.”

At its annual AI World conference in Las Vegas (formerly Cloud World), Oracle has lifted the wraps off three separate projects in collaboration with key hyperscalers in order to better serve customers.

The updates coincide with the general availability of Oracle Autonomous AI Lakehouse – the company’s new all-in-one platform which combines the scalability and flexibility of a data lake with the management and performance benefits of a data warehouse.

Oracle Autonomous AI Lakehouse brings AI and machine learning directly into the database, which means customers can use AI where their data lives instead of having to duplicate it and move it elsewhere.

Oracle is not afraid of multicloud

With the hyperscaler collaboration, it means that customers can now bring Oracle’s AI to their data even if it’s stored in Google Cloud, Azure or AWS, keeping them aligned with data residency and compliance requirements.

Besides offering Oracle Autonomous AI Lakehouse with these partners, Google Cloud has also brought the number of regions up to eight, from five.

Over the course of the next 12 months, Google Cloud has committed to adding a further nine regions, Azure another five, and AWS adding a whole 20 new regions for Oracle customers.

Speaking with TechRadar Pro at the event, Oracle VP of Product Management Nathan Thomas emphasized the company’s commitment to letting customers place Oracle databases in the cloud of their choice, be that on Oracle Cloud Infrastructure or not.

Oracle’s Database@ products are designed to bring low-latency, compliant multicloud deployments to where customers are already running their applications, reducing technical hurdles.

“Our customers want flexibility… we recognize that and we offer a multicloud environment,” Thomas noted.

Oracle Applications VP Rajan Krishnan also shared thoughts on the company’s multicloud strategy with TechRadar Pro at the event, recognizing that customers today pursue not only multicloud, but also multi-AI options where they can bring together the power of different models from different providers.

Krishnan credited Anthropic’s work on the Model Context Protocol (MCP) for enabling broader interoperability, likening it to HTTP’s standardization of the web.

In short, the message at AI World 2025 has been clear in that the future of data is open, flexible and collaborative, and Oracle’s partnerships with companies like these is just proof that they can co-exist in the same space, maintaining rivalry and their own differentiators while also better serving customers’ needs.

You might also like

• We've listed the best cloud hosting providers
• Check out the best AI tools around today
• Here's everything we saw at Oracle AI World 2025

• UK SMEs are stalling ideas because they don’t have enough time, Google report claims
• AI tools could boost productivity by 20%, unlocking £198 billion
• Only one in three use GenAI, one in six use it daily

Nearly two-thirds (59%) of UK SME owners have paused game-changing ideas due to a lack of time, even though many agree innovations could boost their revenue, new research has claimed.

The report from Google Cloud suggests AI tools, like its own Gemini-backed Workspace apps, could raise SME productivity by 20% (the equivalent of a free extra day per week).

With projections that UK SMEs could unlock an extra £198 billion from the productivity gains, Google Cloud has launched a new scheme with the Department for Business and Trade (DBT) and NatWest to get AI into the hands of more SMEs.

Google wants to help UK SMEs get on board with AI

The three have come together to launch a new nationwide tour, visiting Manchester, Leeds, Edinburgh and Cardiff, to show local businesses how AI can improve efficiency and drive innovation.

Google wants to use AI to plug resource gaps, with around two in three (64%) UK SMEs acknowledging that they’d be able to unlock “significant revenue growth” simply with more time.

“There is enormous potential for AI to free up some time for the nation’s business leaders to focus on strategic, innovative tasks that will spur growth for their businesses and the wider economy,” Google EMEA President Debbie Weinstein said.

The scheme comes after the company published its AI Works research report, revealing that only one in three (34%) UK workers use generative AI at work (and only half of those using it daily).

Looking ahead, it seems that a lack of training is preventing SMB workers from adopting AI more broadly, highlighting the need for better resources. Separately in the report, Google noted that a few hours of training was enough to double or triple AI use, giving workers a starting point to go off and do their own learning.

UK Secretary of State for Business and Trade Peter Kyle noted a commitment to partnering with private sector companies, like Google, to deliver training initiatives and other support mechanisms for SMBs.

You might also like

• Use the best online learning platforms to upskill your workers
• Let the best AI writers take the hassle out of administrative work
• SMBs are turning to AI to solve their problems - but many are still cautious to go all-in

• Google Cloud has complained about Microsoft’s unfair licensing practices
• Rival cloud providers are forced to charge for licenses, causing vendor lock-in
• Little has been done to solve the issue in 12 months, Google Cloud says

Google has published a new blog post reminding the world it filed a formal complaint with the European Commission one year ago over Microsoft’s cloud licensing practices, but that nothing still has been done.

The company argues Microsoft penalizes customers who run Windows Server and other software on rival clouds, like its own Google Cloud (GC) as well as AWS and others.

According to the complaint, Microsoft’s licensing rules make it up to 5x more expensive to move workloads to competitors, essentially resulting in vendor lock-in.

Google isn’t dropping its Microsoft complaint

The problem is Microsoft allows existing on-prem licenses to transfer to Azure, but rivals like Google Cloud require customers to buy new licenses, which adds to the cost.

Britain’s Competition and Markets Authority (CMA) has already deemed that Microsoft’s policies restrict cloud choice, but it has not yes imposed any remedies. America’s Federal Trade Commission (FTC) also launched its own antitrust investigation in 2023.

GC VP of Government Affairs and Public Policy Marcus Jadotte and GC Europe Senior Director of Government Affairs and Public Policy Giorgia Abeltino noted: “Restrictive cloud licensing has caused an enormous amount of harm to the global economy over the last year.”

Speaking about “the ongoing risk of inaction,” Jadotte and Abeltino accused Microsoft of acquiring customers at “two or even three times the rate” of competitors.

In a bid to strengthen its case and speed up action, the Google Cloud blog claims governments could unlock up to €1.2 trillion in additional EU GDP by the end of the decade, saving up to €450 billion per year in gained productivity.

Microsoft has already settled similar concerns with European cloud providers, with a company spokesperson telling TechRadar Pro: "Microsoft settled amicably similar concerns raised by European cloud providers, even after Google hoped they would keep litigating. Having failed to persuade European companies, we expect Google similarly will fail to persuade the European Commission."

You might also like

• We’ve listed the best cloud hosting providers
• Keep your data online with the best cloud storage and best cloud backup tools
• Microsoft finally ends EU worries over Teams competition with new pricing shakeup for 365 suites

• The EU Data Act requires companies to offer data transfer ‘at cost’
• Google Cloud removes the cost, AWS and Azure still charge
• Google wants to be seen as an openness champion

Google has scrapped data transfer fees for organizations using multicloud approaches within the European Union and UK.

The change landed just days before the EU Data Act comes into force (on September 12, 2025) – a piece of legislation designed to make it easier to access, share and use data across sectors and platforms.

Although the Act only requires companies to offer data transfer options ‘at cost’, Google went one step further by enabling this for free.

Google makes it free to transfer data in EU/UK

Google Cloud Global Risk and Compliance Senior Director Jeanette Manfra announced the change in a blog post, emphasizing the company’s commitment to interoperability, openness and multicloud strategies.

Manfra stressed that Google was the first company to offer digital sovereignty solutions for European governments, and the first to waive exit fees for customers leaving Google Cloud.

All of this from a company that trails behind AWS (32%) and Microsoft Azure (23%) in terms of market share during Q1 2025 (10% for Google Cloud).

The post also harks back to the original promise of the cloud – an ecosystem that’s “open, elastic and free from artificial lock-ins.”

Looking ahead, Google Cloud customers will need to opt in to the Data Transfer Essentials service to benefit from zero fees. Note that this is designed to work with “in-parallel” workloads only – the same organization, spread across two or more cloud providers.

Although Google’s response to the EU Data Act is late compared with hyperscaler rivals, it’s certainly the most favorable approach. Microsoft began charging fees in the EU in August, and AWS lets EU customers apply for reduced rates in certain use cases.

As such, Google has framed its response as a broader push to challenge market lock-in – a hot topic of discussion among European regulators.

You might also like

• Competition in the cloud and why UK businesses are paying the price
• Looking for an alternative? These are the best cloud hosting providers
• Check out the best cloud storage and best cloud backup tools

• Meta set to partner with Google Cloud for major infrastructure boost
• More Meta data centers are planned soon
• AWS and Microsoft Azure already work with Meta

Meta has agreed to spend at least $10 billion on Google Cloud services in a six-year deal, marking the first major cloud deal between the two giants.

The reports, according to two people familiar with the matter, suggest Meta could use Google Cloud servers and storage for its AI infrastructure, despite investing in its own data centers.

However, CEO Mark Zuckerberg’s grand plans to deliver maximum computing power to every AI researcher means that the company can’t keep up with its own expansion, drawing on the experience of others.

Meta to use Google Cloud for its AI infrastructure

Despite having over 20 data centers in its existing portfolio (and more under construction), Meta needs to draw on this external capacity for its short-term plans, but the move is not unheard of even from Meta.

The social networking giant already uses AWS and Microsoft Azure, so Google just adds to its multicloud strategy.

Google Cloud also previously partnered with Meta by hosting Llama models on Vertex AI.

Google lags way behind Amazon and Microsoft in third place in the cloud market tables, but this didn’t stop it winning a big OpenAI contract earlier in 2025 as the ChatGPT maker looks to diversify beyond its earlier heavy Microsoft investments.

Google Cloud generated $13.6 billion in the company’s second quarter, marking a considerable 32% year-over-year increase compared with the 14% increase seen by the company as a whole.

Looking ahead, Meta’s projected expenses for 2025 could reach $118 billion, with a large chunk dedicated towards the company’s AI efforts.

It saw a healthy 22% rise in revenue during its second quarter, which is roughly how much its 2025 expenses could rise by.

Still, despite Meta’s AI plans, overall headcount at the end of its second quarter was 7% higher than Q2 2024, at nearly 76,000.

Meta did not immediately respond to TechRadar Pro's request for comment. A source close to Google confirmed the reports are correct.

Via CNBC

You might also like

• OpenAI to move to Google Cloud infrastructure to boost ChatGPT computing power
• These are the best AI tools and best AI writers from all companies
• Check out the best cloud computing partners

• Oracle is adding Gemini to OCI, starting with Gemini 2.5
• Further models, including specialized ones like MedLM, will also be added
• OCI recently signed a similar deal to add xAI's Grok, too

Oracle and Google Cloud have announced a deal to make Google's Gemini models available via Oracle Cloud Infrastructure's (OCI) Generative AI service.

Oracle confirmed the service would kick off offering Gemini 2.5, but the company would continue to add further models for text, image, video, speech and music generation as well as specialized models like MedLM for healthcare.

The company explained customers would now be able to use Gemini models – among others – to help them build their own AI agents for coding, automation, research and more.

Oracle Cloud now offers Gemini models

"Oracle plans to make Google’s entire range of Gemini models available via OCI Generative AI service through new integrations with Vertex AI," Oracle added.

Although the models will continue to run on Google's servers, they can be accessed through Oracle's cloud platform with customers paying with Oracle Universal Credits, but the benefits of this partnership will be twofold.

For Oracle, it means the company can expand its offering of models, including third-party options, rather than pushing its own in-house models, making it more attractive for broader audiences. With the deal, Google can also expand its enterprise reach, getting Gemini models into more hands.

"Oracle has been intentional in offering model choice curated for the enterprise, spanning open and proprietary models,” OCI President Clay Magouyrk explained.

The news comes around two months after Oracle struck a similar deal to make xAI's Grok models available to customers, and while many hyperscale cloud providers are battling it out to secure exclusive or preferential AI model partnerships (like Microsoft with OpenAI and AWS with Anthropic), Oracle's multi-vendor strategy could give it a more stable position in the future.

You might also like

• We've listed the best cloud computing platforms
• Check out the best AI tools and the best AI writers from a range of different models
• Oracle signs a mystery huge cloud contract - one which could net it $30bn in revenue

• Google pledges $1 billion in funding for AI education and training across the US
• Eligible students also get 12 months' Google AI Pro, with Gemini 2.5 Pro and more
• The news comes just weeks after Microsoft announced a similar scheme, but at 4x the value

Google has confirmed a commitment to spend $1 billion over the next three years to support AI education and training across the US, targeting colleges, universities and nonprofits.

Alphabet CEO Sundar Pichai explained in a blog post more than 100 colleges and universities have already signed up to its AI for Education Accelerator program, which hopes to make AI and career training free for every college student in America.

Google's training efforts add to its already extensive involvement within education – the company already runs schemes to get Chromebooks into the hands of students in multiple countries and regions.

Google wants to train America's students in AI

In the post, Pichai explained his own personal motivation for the project: "Having regular access to computers in grad school changed my life, and led me on the path to Google. It’s my hope that bringing the best AI tools to college students will open up new worlds for them, too."

The latest announcement sees eligible students across the US, as well as Japan, Indonesia, Korea and Brazil, able to sign up for a year's free Google AI Pro plan with Gemini 2.5 Pro access, Deep Research via Gemini 2.5 Pro, Veo 3, higher limits for Jules, NotebookLM and 2TB of storage.

Boasting about its presence across the education sector, and offering a small dig at Microsoft which previously dominated the area, Pichai noted that more than four in five of the top 100 US universities now use Google Workspace for Education.

However, Google isn't the only company investing in upskilling the current generation of students and our future workforce.

Microsoft President Brad Smith recently shared details of a $4 billion investment in AI and cloud technology for K-12 schools, community and technical colleges, and nonprofits. Other companies in this space, including Meta, OpenAI, Anthropic and Amazon have all pledged their own support, too.

You might also like

• These are the best AI tools and best AI writers
• Check out our roundup of the best online learning platforms
• Microsoft is spending $4 billion to push AI in schools, universities and more

• Google Cloud adds six new AI agents for data scientists, engineers and more
• Advanced analytics will become more accessible with natural-language AI
• A solid data foundation is just as important, but Google can help you migrate

Google Cloud has launched six new AI agent tools to assist data engineers, data scientists, developers and business users realize even more productivity benefits.

Outlining a, "new era where specialized AI agents work autonomously and cooperatively to unlock insights at a scale and speed," Data Cloud Managing Director Yasmeen Ahmad explained the benefits of a "single, unified, AI-native cloud" over siloed tools when it comes to using AI.

Besides new, specialized AI agents, Google Cloud is also launching a series of APIs, tools, and protocols as well as updates to unify data.

Google Cloud launches even more AI agents

The first agent, destined for data engineers, is designed to automate complex data pipelines by allowing engineers to describe tasks and then autonomously building and executing workflows. A separate Spanner Migration Agent will simplify migrating from legacy databases like MySQL to Spanner, eliminating hours of tedious administrative work.

Data scientists will benefit from an agent that automatically performs exploratory data analysis, data cleaning, feature engineering and ML predictions, offering step-by-step reasoning and collaborative feedback, while business users and analysts will get to use two separate agents designed to answer questions about data and interpret code with visualisations and explanations, meaning that non-technical users can perform advanced analytics.

Finally, Gemini CLI GitHub Actions will automate pull requests, tests, reviews and implementation for developers.

"The true potential of the agentic shift is realized when developers not only use existing agents, but also extend and connect them to their own intelligent systems, creating a broader network," Ahmad explained.

With its new agents, Google Cloud hopes to lower the barrier of entry into advanced data analytics, "eras[ing] the line between operational and analytical worlds."

You might also like

• The next big thing in AI is agents, but is your data ready?
• Check out the best AI tools and best AI writers
• We've listed all of the best productivity tools

• Google revenue is up 14%, driven by a 32% growth in Google Cloud revenue
• The cloud division's annual revenue run-rate is now $50 billion
• The Gemini App has 450 million monthly active users

Alphabet has exceeded quarterly revenue expectations by a not-so-insignificant $2.43 billion by posting $96.43 billion in its second quarter 2025, marking a health 14% year-over-year increase, with the company having AI and its cloud services to thank.

Google Cloud revenues rose by 32% in the quarter to $13.6 billion – considerably more than the 12% rise in Google Services revenue, which covers areas like Search, subscriptions, devices, YouTube ads and more.

In a blog post breaking down the earnings, CEO Sundar Pichai revealed Google Cloud's annual revenue run-rate has now passed the $50 billion mark.

Google Cloud boosting Alphabet's overall revenue

"We are seeing significant demand for our comprehensive AI product portfolio," Pichai explained in the post.

"We operate the leading global network of AI optimized data centers and cloud regions."

Pichai added nine million developers have now built with the company's Gemini 2.5 family of models, while 70 million videos have already been generated using Veo 3, which Pichai described as a "viral hit."

Speaking about Google Cloud revenue in particular, Pichai revealed the number of large contracts worth more than $250 million has more than doubled year-over-year. Google Cloud also signed the same number of $1 billion+ deals in the first half of 2025 as it did in the whole of 2024.

Wayfair, Mattel, Target, Capgemini and BBVA were cited among Google Cloud's major clients, with OpenAI also recently confirming that it would use Google Cloud infrastructure to support ChatGPT in the US, the UK, Japan, the Netherlands and Norway.

On the consumer side, the Gemini App now has more than 450 million monthly active users, and daily request have grown by more than 50% since last quarter.

All of this saw stock prices rise by about 3% in after-hours trading, with Google currently ranking as the world's fifth-most valuable company with a market cap of $2.315 trillion, just a few paces behind Amazon.

You might also like

• These are the best AI tools and best AI writers you can use today
• We've listed the best cloud computing providers, including Google Cloud
• OpenAI to move to Google Cloud infrastructure to boost ChatGPT computing power

• Google Cloud will provide OpenAI's infrastructure in a number of countries
• Google also has a cloud deal with OpenAI rival Anthropic
• Then, there are Google's own in-house AI efforts

OpenAI has now added Google Cloud Platform as a cloud provider for ChatGPT and its APIs in a number of countries, according to CNBC reporting.

This comes as the AI giant diversifies its infrastructure to meet growing demand – the company also famously uses Microsoft, as well as Oracle and CoreWeave.

Once the company's exclusive cloud provider, Microsoft has lost exclusivity as the relationship between the two companies evolved. Today, Redmond still has first right of refusal, but it's no longer the sole vendor.

OpenAI adds Google to its list of cloud providers

OpenAI's use of Google Cloud Platform is a significant win for the company, whose cloud business is much smaller than AWS and Microsoft Azure even though it's the third-largest cloud provider globally. Its infrastructure will be used in the US, the UK, Japan, the Netherlands, and Norway.

On the other side of the deal, OpenAI isn't Google's only AI customer. The tech giant already has a partnership with Anthropic – a rival AI company founded by former OpenAI execs.

OpenAI CEO Sam Altman has previously cited a shortage of GPUs as a motivation to not only increase, but broaden the range of cloud services.

According to Reuters, the deal with Google could have been on the cards for a number of months, but it wasn't until the exclusive partnerships with Microsoft ended that discussions could progress.

For OpenAI, diversifying suppliers reduces single-vendor lock-in, ultimately giving the company more leverage to negotiate on pricing and to access the latest chips.

However, the deal raises questions about how Google and its rivals will allocate resources, each offering their own alternative to ChatGPT. Looking ahead, OpenAI is on track to finalize accelerator designs soon, which could further reduce its reliance on third parties.

You might also like

• These are the best AI tools and best AI writers
• Access the most powerful chips with the best cloud computing providers
• OpenAI just signed a huge deal with the US government - and it could change the future of AI as we know it

Despite being an increasingly common presence in homes and businesses across the world, the wider benefits of AI are still being viewed with suspicion by some.

In the enterprise world, the opportunities are perhaps much clearer, and Google Cloud is among the firms looking to show off these benefits to all.

At the recent Google Cloud Summit London, I got to speak to Maureen Costello, Vice President UK, Ireland, and Sub-Saharan Africa, Google Cloud, to find out how the company sees the current state of the market - and what can be done going forward to encourage these opportunities even further.

Leaning into AI

“I think we’re starting to see, across all industries, AI being used in a very ‘everyday’ way for customers," Costello notes, ahead of speaking at her keynote at the event.

“We’ve seen customers are really leaning into AI…moving from experimentation, and really leaning into using AI as a proactive collaborator.”

A large part of this is AI agents, which offer a lot of promise in streamlining processes and freeing up time and resources for over-stretched human workers.

The company unveiled a host of agentic AI tools and services at its Google Cloud Next 25 event earlier this year, including several upgrades to its Agentspace platform to make agent discovery and adoption easier.

Among the announcements were a new Agent Gallery to find and deploy new agents quickly, as well as a tool to create agents with the new no-code Agent Designer platform, and Agent2Agent, a new interoperability protocol which will allow AI agents to communicate with each other.

"Customers are really leaning into the agentic AI era and figuring out what this means for their business," Costello notes, "and they're looking to us to provide them with the tools and the technology around that…(so) we really provide them with the full toolkit around AI."

Top of the agenda for many at Google Cloud Summit London was a new partnership between the company and the UK Government, which will see up to 100,000 civil servants receive training on using AI effectively.

But the event was also a chance for Google Cloud to show off the wealth of customers it has across the region, and how they are using its tools (and AI) to grow and prosper.

“It’s a strong partnership with the UK government and UK businesses,” Costello notes, “as AI is changing the way that work gets done, we want to make sure that we’re also leaning in to work with both businesses and the government in terms of uplifting the workforce capability.”

“We are investing heavily in helping the UK put in place that digital foundation...it's making sure people are able to focus on what they really want to be doing at work, or as jobs change, making sure they have the capability to change that…we see it as getting the collaborator into the workforce.”

This is not just talk either - Costello notes Google’s own developers and engineers are using its AI-enabled Code Assist tool internally to help them write code, a boost she says has helped change how these workers are doing their jobs - as she notes, it’s collaboration, not opposition.

Open, trusted, and secure

With AI dominating the news agenda, it can be tricky for businesses to choose which provider to go with - but for Costello, there's unsurprisingly only one choice.

“Our approach to cloud is fundamentally different - it’s built on a foundation of being open, trusted, and secure,” she notes, “we want to give our customers the freedom to innovate without locking them in to particular vendors - but also being able to protect their data and make sure that they’re doing that with the very best security on the planet.”

“It’s really important to us that customers are able to have freedom through the open cloud, so our commitment to openness is around open source technologies such as Kubernetes - that means businesses can run our cloud, but they can also run on their own data centers and other clouds, because we want our customers to choose us because we’re the best cloud, rather than them feeling that they’re forced into a commercial contractual reason.”

“From a security point of view, we believe that we have the best security on the planet - we protect from a consumer point of view, we have so many people using Gmail, YouTube every day - and this is the security we bring into our enterprise products," Costello adds, "and we continue to invest - if you look at our current acquisition of Wiz, that will allow us to be able to help our customers, protect their full environment regardless of which cloud they’re using, and bring the intelligence that Mandiant gives us as well.”

“So we’re really looking to help our customers protect their full estate, regardless of whether they’re using us, or other clouds - we recognise customers want to have control of their data, and access to their data - so for us, it’s all around making sure the relationship we continue to have with our customers is about openness, trust, and security.”

So going forward, the message from Google Cloud seems crystal clear - AI is here to stay, and your business can realise much more potential the more you embrace it.

• These are the best cloud hosting options around today

• Google reportedly set to offer the US Government discounts on cloud services
• It's already discounted its business software earlier this year
• Other hyperscalers are also discounting their cloud services

Google has become the latest in a growing list of tech companies to offer heavily discounted services to the US government, this time for its cloud services.

As reported by the Financial Times, the deal, which could be finalized within a matter of weeks, reflects an emerging trend across the sector, with tech firms all bidding to win big government contracts with attractive discounts.

The trend comes in response to President Trump's efforts to centralize government spending – by buying in bulk, Trump apparently hoped to secure bigger discounts, and that's exactly the response he has seen.

US government unlocks major Google Cloud discounts

Although Google and the General Services Administration (GSA) are yet to share any details regarding the reported discounts, we know that other technology firms have also shared major discounts recently.

This includes Oracle, which is offering the US government a 75% discount on license-based software and "substantial" discounts on other cloud services.

As Oracle and Google are both offering discounts on cloud services, and they're likely to be joined by others, it's unclear whether the White House is set to choose one supplier, or if it will spread services across different operators.

Microsoft Azure and Amazon Web Services are also expected to offer comparable discounts to maintain competition against their hyperscaler rivals.

Although Google's latest discounts have not been confirmed, the company did offer a 71% discount on business apps earlier in 2025, which was set to save the US government up to $2 billion.

TechRadar Pro has asked Google to share any details on ongoing negotiations with the GSA, but we did not receive an immediate response. Any updates will be posted here in due course.

You might also like

• Cut costs with the best free office software
• Check out the best cloud storage and best cloud backup tools
• UK Technology Secretary tells tech giants to “bring their best tech, at the best price”

• Technology Secretary calls for more investment in UK from tech giants
• Peter Kyle hails potential of UK businesses
• Kyle hits back on those who criticise his meeting with tech firms - as that's his job

The UK Government’s Technology Secretary has called for global tech giants to continue bringing their latest innovations, and more investment to the country.

Speaking at Google Cloud’s Summit London event, Peter Kyle declared, “my message to the big technology companies is clear: bring us your best ideas, your best tech at your best price, and you will get access to the biggest client in the country.”

“In return, you’ll get access to the biggest client in the country, one that will be increasingly intelligent and increasingly digital.”

Goodbye "ball and chain" tech

Kyle also hit out at those who criticised his meetings with technology firms, noting, “that is my job”.

He mentioned a report from The Guardian which claimed he had 28 meetings with people in or close to tech sector in his six months in the role, including major firms such as Amazon, Google, Microsoft, and Meta - 70% more than his predecessor.

“Yes I’ve met tech companies”, he declared. “That’s how you deliver value for the public. That’s how you unlock innovation”.

Kyle was speaking as the UK Government announced a new partnership with Google Cloud which has the aim of helping drag the former away from the reliance on “ball and chain” legacy technology.

He noted more than one in four public sector tech systems are estimated to run on legacy tech - a figure which jumps to as high as 70% in some police forces and NHS trusts, sometimes on contracts signed decades ago.

Overall, the partnership could see Google invest hundreds of millions of pounds in Britain’s public sector technology, Kyle noted, "helping to deliver my ambition to bring the public services people use every day, drag it into the 21st century."

The move would also help lessen the burden on the taxpayer, who often has to foot the bill for the estimated £21 billion spent on buying the same technology time and time again due to outdated, locked-in contracts.

Kyle referenced several other significant recent announcements, including the recently-released GOV.UK app, as well as the National Digital Marketplace, a new procurement platform for the public sector.

"Britain will be using technology in more areas and more than ever before," he declared, "We know that tools using the same technology are capable of transforming Whitehall itself, the NHS, and other essential services that millions of people across our country rely on (and) with more hands-on support, I can’t wait to see what our two teams deliver together."

You might also like

• Delivering the AI citizen experience – the future of government
• Can the UK be a big data leader in the military?
• These are the best AI tools around today

• UK Government and Google Cloud sign new partnership
• Google Cloud looks to help modernize aging UK Government tech stack
• 100,000 civil servants set to receive more training on AI and other tech

Google Cloud has announced a major partnership with the UK Government which will help the country rid itself of aging technology systems.

Announced at the company's Google Cloud Summit London, the agreement will look to help the UK Government update a host of legacy tech, some of which is decades old, to unlock up to $45 billion in efficiency savings.

The partnership will look to modernize commonly-used services from the NHS, along with local council tools, covering everything from bin collection to tax returns.

So long "ball and chain"

Speaking at the Summit, Technology Secretary Peter Kyle noted the deal will look to provide a major upgrade across public sector services in the UK, where more than one in four systems are estimated to run on legacy tech. This jumps to as high as 70% in some police forces and NHS trusts, sometimes on contracts signed decades ago.

He noted how in some areas, the technology being used is decades-old, leaving these organizations at risk of cyberattack or outages, and meaning workers lack the efficiency and productivity gains offered by newer tech.

Google Cloud hopes its assistance will help Government bodies ditch these "ball and chain" tech contracts, where providers have often made it too difficult to transfer data to modern systems, with some companies hiking the prices while the taxpayer is locked into contracts.

"With more hands-on support, I can't wait to see what our two teams deliver together," Kyle added.

Looking forward, Google Cloud will also help the UK Government explore the usage of technology such as quantum computing within the public sector, as well as investigating the development of a single platform to monitor and keep the public sector safe from the latest cybersecurity threats.

“Google Cloud is proud to be a strategic partner of the UK Government as they embark on this ambitious digital transformation," noted Tara Brady, President, Google Cloud EMEA.

"As the country evolves, the flexibility of Google Cloud will help the government lead the AI era."

Google Cloud will also be helping to boost workers within the UK Government, with a new training programme seeing 100,000 civil servants get extra help with AI and other new technology by 2030.

Elsewhere, Google Cloud is running a 12 week equity-free accelerator program for Seed to Series AI startups based in the UK, in partnership with Tech London Advocates.

This follows benefits announced earlier this year as part of the “Gemini for UK” initiative, with Google Cloud providing up to £280,000 in cloud credits for eligible UK-based AI startups.

The company sys that since 2023, over 60% of genAI UK startups are Google Cloud customers, and it is looking to help some of these firms even further, securing spaces for four top UK AI startups on the Europe-wide Google for Startups accelerator, who will be joining 15 top AI startups across Europe to get Google expertise to accelerate growth.

You might also like

• "AI is simplifying technology itself" - Google Cloud CEO outlines AI hopes
• “The gloves are off when it comes to innovation” - Google Workspace head
• These are the best cloud hosting options around today

• Google Cloud's API service ere to blame for widespread outage
• Most regions were back online in 40 minutes, but some took even longer
• The company has promised to protect against future outages and improve communication

Following Google Cloud's recent widespread outage, which took sites like Spotify, Cloudflare and Discord offline, the company released its detailed report sharing exactly why it failed customers.

The company says the root cause was a code issue in Service Control – part of the company's API management and policy checking system.

Specifically, invalid automated quota update and a lack of proper error handling triggered a global crash loop, with 503 errors seen across not only Google Cloud services, but services using its APIs.

Google Cloud outage caused by API issue

The outage affected the Google Cloud infrastructure, as well as other popular Google Workspace apps like Drive, Docs, Gmail and Calendar. However, third-party sites accessing Google Cloud's API, including popular music streaming platform Spotify which boasts of 678 users, as well as some Cloudflare services, were also affected.

"On May 29, 2025, a new feature was added to Service Control for additional quota policy checks," the company wrote in its incident report. "The issue with this change was that it did not have appropriate error handling nor was it feature flag protected."

Google Cloud boasted that its Site Reliability Engineering team had started triaging the incident within two minutes, having identified the root cause within 10 minutes. "The red-button [to disable the serving path] was ready to roll out ~25 minutes from the start of the incident," Google said, with the rollout complete within 40 minutes.

Although smaller regions recovered relatively quickly, larger regions like us-central-1 took longer to come back online – around two hours and 40 minutes in the case of this particular region.

In its mini incident report issues on the day of the outage, Google Cloud promised to "do better." Its more detailed report promises the usual responses going forward, such as improving static analysis and testing practices, auditing and modularizing Service Control’s architecture to contain future incidents, but the company has also pledged to "improve [its] external communications" to better inform customers, ensuring that its communications infrastructure remains online even during such outages in the future.

You might also like

• These are the best cloud hosting providers
• We've rounded up all of the best web hosting providers
• Google Cloud has big plans to take the pain out of adopting AI agents in your business

• Google Cloud outage affected 40 locations, and even Cloudflare
• Spotify, Discord and more went down, as were some Workspace apps
• Google blamed a dodgy update for the outage, promises to "do better"

A number of popular websites went offline yesterday after Google Cloud experienced a major outage which saw users unable to access the likes of Spotify, Discord and Cloudflare for some time.

Impacting over 40 locations and 26 services, the outage saw systems go down at 11:46am PT on Thursday, June 12, but by 12:30pm PT, Google reported the issue was mostly resolved, except for the us-central1 (Iowa) Region.

The company claimed it had identified the cause of the issue and that engineering teams were working on a recovery, but some users continued to experience issues at 12:41pm PT.

Google Cloud outage takes down popular sites

Although Google Cloud stated the outage started at 11:46am PT, Cloudflare also experienced disruptions from 11:19am PT, blaming Google Cloud for the issues.

Among the popular sites affected were Spotify, Discord, Snapchat, Character.AI, Cursor and Replit. Downdetector also shows spikes in the number of complaints from users trying to access some Google services, like Google Maps and YouTube.

According to the outage tracker, some Workspace apps like Drive, Docs and Meet were also affected by the outage.

"We are deeply sorry for the impact to all of our users and their customers that this service disruption/outage caused. Businesses large and small trust Google Cloud with your workloads and we will do better," the company wrote in its mini incident report.

The report blames an invalid automated quota update to Google Cloud's API management system for the outage, claiming almost full recovery within two hours. However, some delays to restoration continued: "Several products had moderate residual impact (e.g. backlogs) for up to an hour after the primary issue was mitigated and a small number recovering after that."

A full log of the incident is available on Google Cloud's status website, and the company has promised a full report to follow.

You might also like

• We've listed the best web hosting providers
• Check out our roundup of the best cloud hosting providers
• ChatGPT is down again – here's everything we know about its latest massive outage

If you experienced issues with a single part or various parts of Google’s massive operation, or with playing your favorite songs or podcasts on Spotify, you weren't alone earlier today.

Starting at 2 PM ET in New York City, reported issues on Down Detector began spiking for Google, Google Cloud, and Spotify, with Google confirming that issues were impacting its various services as of 3:01 PM ET.

Spotify normally comments on issues via the @SpotifyStatus account on X (formerly Twitter), but remained completely silent as users around the world began to have issues. The TechRadar team utilizes various Google services, primarily G-Suite with Docs, but didn't encounter any issues during this outage.

Other services on Down Detector began spiking with reports, including Snapchat and Discord, as well as cloud data provider Cloudflare, which later confirmed on its status page that a key third-party was having an outage. It was later confirmed that the culprit was Google Cloud.

In total, the outages impacting Google, Spotify, and Cloudflare (in the form of that third-party) lasted for several hours, but have since returned, with Google Cloud's status page showing the incident complete as of 9:30PM ET. You can read our live reporting on the outages affecting Google and Spotify below.

A Quick Look at Down Detector’s homepage as of 3:25 PM ET shows that an extensive range of services are experiencing issues, with some leveling off or dropping, with Google, Google Cloud, and Spotify at the forefront, alongside Discord.

Google and Google Cloud both began spiking in the 2 PM ET hour, reaching over 10,000 reports, while Spotify also started in that hour but is currently at over 44,000 reports.

Google is investigating the issue

Google's status page currently lists an active 'Service Disruption' as of 3:01PM ET that is impacting a number of services – it's a good note, though, as teams at the company are investigating.

"We're investigating reports of an issue with Gmail, Google Calendar, Google Chat, Google Cloud Search, Google Docs, Google Drive, Google Meet, Google Tasks, and Google Voice. We will provide more information shortly.

Multiple Workspace products beginning on Thursday, at 2025-06-12 10:58 PDT may be experiencing service issues.

Our engineers are currently investigating the issue.

We apologize to all who are affected by the disruption."

It's impacting several services from the entire G-Suite, including Gmail. Still, Google doesn't clarify if a specific user set is experiencing the issue. As of now, I can access my personal and work Google accounts without problem. However, individuals in the comments on Down Detector are reporting issues with Messages, Google Cloud, and Google Voice.

Spotify has still not issued a statement, although reported issues on Down Detector continue to grow, now standing at over 45,000 reports as of 3:22 PM ET. We have also reached out to the streaming giant to request a comment.

It does seem that a majority of these issues started shortly after the 2 PM ET hour and are now stretching to over an hour of disruptions.

At the same time that these issues with Google and Spotify began emerging, Cloudflare is dealing with its own problems, according to the company's status page.

The latest update, as of 3:12 PM ET, from Cloudfalre notes that its services are starting to recover, but issues are still present.

"We are starting to see services recover. We still expect to see intermittent errors across the impacted services as systems handle retried and caches are filled."

Considering that Discord and Snapchat are also experiencing a spike in reports, these issues may well be related to the problems affecting Cloudflare.

Google's making progress, says everything but Meet is fixed

Google is making some progress, at least according to the latest update on its Workspace status page, posted at 3:30 PM ET.

It reads: "All product impacts except Google Meet have recovered.

Google engineers continue to work on full mitigation."

This is good news for folks in the Google ecosystem, as it appears everything but Google Meet is back up. And even reports for Google's calling platform are starting to drop on Down Detector, now sitting at 1,854 reports as of 3:31 PM ET.

Snapchat and Discord are both seeing reported issues spikes

Alongside issues impacting Google, Google Cloud, and Spotify, reported issues with Snapchat and Discord are both elevated on Down Detector.

Reported issues with Discord are currently sitting at 6,683 as of 3:35PM ET, but did spike to over 10,000 at 2:20PM ET – the same hour where Google and Spotify saw major increases.

Snapchat is currently on the rise, with over 6,693 reported issues with the platform. For what it's worth, I can open the app on my iPhone, but I'm unable to load stories or the main page.

Google gives the all clear on Workspace issues

The Google Workspace status page now effectively says we're all clear. In a post that went live 3:53PM ET, Google says all the issues are resolved.

"The problem with Gmail, Google Calendar, Google Chat, Google Cloud Search, Google Docs, Google Drive, Google Meet, Google Tasks, and Google Voice has been resolved. We apologize for the inconvenience and thank you for your patience and continued support."

Reports on Down Detector have slowed for both and have been on the decline, so it appears that most services are coming back or already back. Again, this does line up with Cloudflare's reported issues, and those are starting to recover.

Cloudflare is having some critical issues

Cloudflare has posted a new update as of 3:57 PM ET on its own status page, detailing a bit more about what is going on and the potential impact here. You can see the statement in full below, but Cloudflare’s critical Workers KV service went offline due to a separate outage hitting a key third-party service.

A Worker KV is essentially a flow or automation that moves requests throughout Cloudflare's vast network, mainly starting or ending in storage libraries.

Furthermore, Cloudflare acknowledges that it's aware of the significant impact this is causing and is working to resolve the issue as soon as possible with all hands on deck.

Cloudflare’s critical Workers KV service went offline due to an outage of a 3rd party service that is a key dependency. As a result, certain Cloudflare products that rely on KV service to store and disseminate information are unavailable including:

Access

WARP

Browser Isolation

Browser Rendering

Durable Objects (SQLite backed Durable Objects only)

Workers KV

Realtime

Workers AI

Stream

Parts of the Cloudflare dashboard

Turnstile

AI Gateway

AutoRAG

Cloudflare engineers are working to restore services immediately. We are aware of the deep impact this outage has caused and are working with all hands on deck to restore all services as quickly as possible.

Google Cloud is still having issues

While Google says the issues affecting Workspace have been resolved, but the Google Cloud dashboard continues to display issues.

There are still "Multiple GCP products are experiencing Service issues," in fact, it's over 39 products including API Gateway, Agent Assist, and AlloyDB for PostgreSQ across the globe. This is the latest update from Google Cloud as of 3:56PM ET, you can see the status page here.

Impacted services are starting to recover

Google, Google Cloud, Spotify, Snapchat, and Discord, among other services that saw an increase in reported issues on Down Detector, are all starting to show a decline, and that's a good thing.

It's been roughly two and a half hours since we started seeing a spike in the 2PM hour for services like Google and Spotify, with the latter seeing over 44,000 reported issues. While Spotify has yet to provide any comment, Google Workspace and Google Cloud have been updating status dashboards. The former states that things are back to normal, while the latter continues to show some impacted services.

Down Detector is looking a lot better now, though reported outages do remain for all these platforms.

Cloudflare is still working to bring all of its impacted services back online, with the last update on its dashboard at 3:57 PM ET. That concluded with, "We are aware of the deep impact this outage has caused and are working with all hands on deck to restore all services as quickly as possible."

Cloudflare say it's services are 'recovering quickly' around the globe

In line with impacted services reported outages dropping on Down Detector, Cloudflare says its services are 'recovering quickly' across the globe in an update on its status page as of 4:32PM ET. It's expecting a 'steady drop' in services impacted and 'further recovery' in the next few minutes.

That's good news and likely means that Google, Spotify, and other services will be back online for you soon, if not already.

"Cloudflare services are recovering quickly around the globe. WARP and Turnstile are operational, though a small residual impact remains and we’re working to eliminate it. The core KV service is restored, bringing dependent products back online. We expect further recovery over the next few minutes and a steady drop in impact."

Now, Cloudflare says it's back to fully operational in an update that was posted just before the top of the hour at 4:57PM ET.

It reads in full: "All Cloudflare services have been restored and are now fully operational. We are moving the incident to Monitoring while we watch platform metrics to confirm sustained stability."

It's an excellent update for those who have felt the impact of this outage, and hopefully, any issues you've experienced have been resolved. While many services were impacted today, alongside this Cloudflare outage, Down Detector is looking a lot better with declines.

Google Workspace's status page indicates that the incident is resolved, while Google Cloud's status page still displays an active incident worldwide.

Furthermore, although Spotify didn't confirm an issue, the brand's care account is responding to a few users, recommending a restart of the app if they're unable to use the service.

Cloudflare comments

Cloudflare was already clear in its last status post that key "3rd party service" was the cause for the issues, and now a spokesperson has confirmed that to TechRadar.

The statement reads, "This is a Google Cloud outage. A limited number of services at Cloudflare use Google Cloud and were impacted. We expect them to come back shortly. The core Cloudflare services were not impacted."

Essentially, saying that Cloudflare's own operations and services didn't have an issue. Noting that it was a Google Cloud outage and that some of Cloudflare uses that system, so they were impacted. Additionally, Google Cloud's status page still shows an active incident.

Spotify's looking into things

While Down Detector reports have dropped considerably for Spotify, the streaming service has made its first comment on the issue on its community blog under a section titled "Ongoing Issues."

It was posted about 40 minutes ago, and reads: "Hey folks! We've been getting reports about issues with login, playback and the Support site. This is now being looked into."

It's unclear whether there are more issues with Spotify, as Down Detector currently reports 1,648 issues as of 5:36PM ET.

Google Cloud says the issues are resolved

As of 9:27PM ET, the Google Cloud status monitoring page says that the incident that impacted a vast number of services has been resolved. The full statement reads:

"Vertex AI Online Prediction is full recovered as of 18:18 PDT.

All the services are fully recovered from the service issue

We will publish analysis of this incident once we have completed our internal investigation.

We thank you for your patience while we worked on resolving the issue."

It's good news that Google Cloud has recovered, and most reports for other services on Down Detector have all declined.

Seemingly, and according to Cloudflare, Google Cloud services were the primary, or at least a significant portion of the issue that caused other services to go down today, June 12, 2025. That includes Spotify and other Google services, such as G Suite and Workspace, as well as Discord and Snapchat.

• OpenAI brings Google Cloud on board for cloud computing infrastructure
• Microsoft is no longer OpenAI's sole provider after a shakeup
• Google shares are up, Microsoft shares down

Despite having a long history with Microsoft, OpenAI is believed to have signed a deal with Google Cloud to help it meet rising demands for cloud computing services.

Reuters reports the deal was finalized in May 2025 after months of negotiating, citing an unnamed source familiar with the matter.

The finer details of its Google Cloud deal are unknown, but the move reflects OpenAI's effort to reduce its reliance on a single vendor – in this case, Microsoft – having recently announced its own $500 billion Stargate Project earlier in 2025.

OpenAI agreement with Google Cloud

Although Microsoft has fuelled much of OpenAI's growth with cloud infrastructure throughout a period of rapid expansion, the company stopped being its sole partner in March 2025, when OpenAI brought CoreWeave on board in a deal worth around $12 billion.

Now, Google Cloud will also provide resources to help OpenAI train and run its AI models.

During its first quarter of 2025, Google Cloud generated $12.3 billion in revenue, or nearly 14% of Google's total revenue. A deal with OpenAI could significantly increase the cloud division's revenue.

However, ChatGPT's success has allowed OpenAI to explore in-house opportunities.

Besides the Stargate Project, in collaboration with SoftBank and Oracle, OpenAI is also working on its own AI chip, which could spell disaster for Nvidia, which has once again taken prime position as the world's most valuable company, with a market cap of $3.510 trillion.

Sceptics are also uncertain about how Google Cloud's deal with OpenAI could impact its existing operations and customers, with the cloud hyperscaler already struggling to meet demand on occasions.

Google shares rose 2.1% after the news, with Microsoft shares dipping 0.6%, pulling it into second position with a market cap of $3.500 trillion.

You might also like

• We've listed the best AI tools
• Get a hand crafting your best work with the best AI writers
• "Just do it": Sam Altman says businesses looking to adopt AI should go ahead – or risk being left behind

• Google Cloud operates across 42 regions and 127 zones, supported by 33 subsea cables
• The US government trusts Google Cloud Air-Gapped with top-secret information
• Europe has a number of local tools, and Germany is next

Amid ongoing US-China tensions over tech dominance, Google has recognized that some countries are concerned about the US' tech dominance, thus it has announced some updates to its sovereign cloud services to ensure optimal privacy.

In a blog post, Google Cloud President of Customer Experience Hayete Gallot boasted that the company now has more than 42 cloud regions, 127 zones, 202 network edge locations and 33 subsea cable investments supporting its global cloud services.

Once niche and limited to highly regulated sectors like defense and intelligence, sovereignty is now a mainstream concern for many sectors and most governments, and Google has responded with a handful of updates.

Google upgrades its sovereign cloud portfolio for customers

Key to the announcement is Google Cloud's air-gapped cloud service, which operates as a standalone cloud with no external network connectivity requirements. It's designed specially for sectors with strict data residency and security, and is build on open-source components which Google believes can enhance resilience and therefore business continuity.

The company even got authorization to store US government Top Secret and Secret-level data on Google Cloud Air-Gapped in 2024 – a sign of its strong privacy credentials.

The second of Google's three key announcements, Google Cloud Dedicated, is a partner-operated regional platform that meets local sovereignty standards. The company already partnered with Thales in 2021 to make its first-ever Trusted Cloud by S3NS for Europe, but now it's confirmed it'll be targeting Germany next.

Finally, Google Cloud Data Boundary gives users further control over where their data is stored and processed, with enhanced protection via external key management and confidential computing. For example, Workspace users can select to restrict processing to the US or EU, while also choosing local data storage countries.

A Mandiant-backed tool called User Data Shield is being added to this to ensure the ongoing verification of sovereignty postures.

Gallot expressed a commitment to "empowering organizations globally to navigate the complex landscape of digital sovereignty with confidence," while also noting how advanced security measure like a zero-trust posture, post-quantum cryptography and AI-powered defenses form part of the broader strategy.

You might also like

• We've listed the best cloud hosting providers
• Keep your data online with the best cloud storage and best cloud backup tools
• Google Cloud unveils new AI Protection security tools, no matter which model you use

• Google claims its Ironwood TPU is 24x faster than the El Capitan supercomputer
• An analyst says Google's performance comparison is “perfectly silly”
• Comparing AI systems and HPC machines is fine, but they serve different purposes

At the recent Google Cloud Next 2025 event, the tech giant claimed that its new Ironwood TPU v7p pod is 24 times faster than El Capitan, the exascale-class supercomputer at Lawrence Livermore National Laboratory.

But Timothy Prickett Morgan of TheNextPlatform has dismissed the claim.

"Google is comparing the sustained performance of El Capitan with 44,544 AMD ‘Antares-A’ Instinct MI300A hybrid CPU-GPU compute engines running the High Performance LINPACK (HPL) benchmark at 64-bit floating point precision against the theoretical peak performance of an Ironwood pod with 9,216 of the TPU v7p compute engines," he wrote. "This is a perfectly silly comparison, and Google’s top brass not only should know better, but does."

24X the performance of El Capitan? Nope!

Prickett Morgan argues that while such comparisons are valid between AI systems and HPC machines, the two systems serve different purposes - El Capitan is optimized for high-precision simulations; the Ironwood pod is tailored to low-precision AI inference and training.

What matters, he adds, is not just peak performance but cost. "High performance has to have the lowest cost possible, and no one gets better deals on HPC gear than the US government’s Department of Energy."

Estimates from TheNextPlatform claim the Ironwood pod delivers 21.26 exaflops of FP16 and 42.52 exaflops of FP8 performance, costs $445 million to build and $1.1 billion to rent over three years. That results in a cost per teraflops of $21 (build) or $52 (rental).

Meanwhile, El Capitan delivers 43.68 FP16 exaflops and 87.36 FP8 exaflops at a build cost of $600 million, or $14 per teraflops.

"El Capitan has 2.05X more performance at FP16 and FP8 resolution than an Ironwood pod at peak theoretical performance," Prickett Morgan notes. "The Ironwood pod does not have 24X the performance of El Capitan."

He adds: "HPL-MxP uses a bunch of mixed precision calculations to converge to the same result as all-FP64 math on the HPL test, and these days delivers around an order of magnitude effective performance boost."

The article also includes a comprehensive table (below) comparing top-end AI and HPC systems on performance, memory, storage, and cost-efficiency. While Google’s TPU pods remain competitive, Prickett Morgan maintains that, from a cost/performance standpoint, El Capitan still holds a clear advantage.

"This comparison is not perfect, we realize," he admits. "All estimates are shown in bold red italics, and we have question marks where we are not able to make an estimate at this time."

You might also like

• AMD-powered El Capitan becomes officially the world’s fastest supercomputer
• Tech editor takes close snaps of world's fastest supercomputer
• Google Cloud unveils Ironwood, its 7th Gen TPU to help boost AI performance and inference

With AI seemingly set to transform how we work at all levels, getting the most when it comes to productivity and efficiency is a key consideration for organizations everywhere.

The recent Google Cloud Next 25 event was unsurprisingly chock-full of announcements on how AI is set to transform the way we work, but how realistic are these aims?

I got to speak to Kristina Behr, VP Product Management, Collaboration Apps, Google Workspace, to find out just how close we really are to the company’s goals when it comes to AI working.

Time to innovate

“I'm really excited about it as a technologist,” Behr tells me, “it just feels like we can do so much by thinking about what is causing people wasted time or drudgery in their jobs, and how can we go about fixing that?”

We’re speaking not long after the main Google Cloud Next 25 keynote, where a host of improvements were announced for Google Workspace, including new audio tools in Google Docs which can create listenable versions of your document, as well as upgraded analysis tools for Google Sheets to help get the most out of your data.

“I really feel the gloves are off when it comes to innovation,” Behr notes, “instead of us designing a separate add-on package with AI...it's like, let's just do this thing, create and really transform (the idea of) what does Google Docs look like, what does Sheets look like for the future, with AI, and really make those two things inextricably linked.”

At a previous Q&A session with the media, Behr had spoken of Google’s aim of “doing things differently” when it comes to enterprise software, and I ask her to expand on what she means.

“We have really great competitors in this space, with probably 10x the amount of people working on these products,” she notes, “so instead of doing things like they would do it, (we think about) how can we do it in new ways, embracing this new technology?”

There are a lot of “unique advantages” that help set Google apart, she adds, including AI-accelerated hardware, which means, “we can deliver a lot of these things with efficiency and scale that other people cannot,” as well as its own proprietary LLMs, such as Google DeepMind, and the applications which sit on top of that.

“Because we can optimize across that stack, I think this is a different way that we can approach this space than some of our competitors - and you’ll see that in both the speed at which we launch new things, but also the quality at which we can optimize that.”

"Game-changer"

Behr had also declared Google Workspace now boasts over three billion users worldwide - with such a huge customer base, I ask her how the company identifies what is important to focus on and prioritize next?

“One of the things we think a lot about is, how do we design the most useful products possible,” she notes, adding a lot of work goes into just how customers are using their products.

“We have decades of knowledge on how people do productivity - and that also gives us a significant advantage on…how do we modernize these workflows.”

Unsurprisingly, AI is once again the answer - Behr says it can hit on the users' needs more precisely, becoming a “game-changer” in the company’s design philosophy.

“Now, everyone is an AI product manager,” she says, “we really go through and think about making our tools as helpful and as useful as possible.”

I ask Behr how the human element and input can remain central in this new AI age.

“I hope this is a false choice,” she laughs, highlighting her own personal usage of Gemini when writing a document to stay in the flow of work, or getting feedback on her output, “I think there’s a way that we can use AI to keep people in the zone, versus being a distraction.”

Personalization is another key selling point for AI tools, Behr notes, as some LLMs can sound quite generic.

“A lot of the work that we’re doing is making sure (AI) is grounded in your data so that it sounds a little bit more like you,” she says, “so if I’m asking for content suggestions, it has access to all the files I have access to, so it’s pulling and mimicking the language, so it’s more personalized to me.”

Authenticity is the key, she adds, as the platform learns over time, but it’s important that “the human is still in the loop.”

“The list of things that we want to deliver for our customers is quite vast,” she notes, “(but) we let the customers be our guide…we try and think about what are the tasks which are common across multiple different segments, different industries - and then, how do we solve them?”

With businesses of all sizes facing a range of cybersecurity threats on a daily basis, the need for a strong and intelligent threat protection offering has never been more crucial.

At its recent Google Cloud Next 25 event, the company was understandably keen to tout its cybersecurity leadership, unveiling a range of new tools and services, with AI unsurprisingly playing a major supporting role.

To find out more on what threats businesses should really be worried about, and to learn more about Google’s own security priorities, I spoke to Sandra Joyce, Vice President of Google Threat Intelligence Group, at the event.

Cyber superpower

Cyber threats can now originate from any country, but Joyce highlights the sheer amount of possible risks coming from “the big four” - Russia, Iran, North Korea, and most notably - China.

China is, “probably the biggest (threat)...they’re getting so hard to detect,” Joyce declares. “They have, I would say, completed their journey to cyber superpower status.”

“There’s likely a capability we haven’t seen, but certainly espionage is first and foremost China’s big lever to pull,” Joyce explained.

“Their capabilities are increasing in ways that are very concerning,” she says, highlighting the recent Salt Typhoon attacks against critical US infrastructure as evidence of the nation’s growing strength in cyber operations.

"We're looking at a major increase in capability,” Joyce says, “they’re leveraging what we’re calling the visibility gap and concentrating their efforts on those areas where endpoint detection and response solutions (EDRs) don’t traditionally operate, like firewalls and edge devices.”

Joyce notes that her team used to be able to detect Chinese threat actors “pretty easily” via the infrastructure being used - however the criminals have now switched to using rented hardware, which is refreshed every 30 days and operated in small offices.

Given the scale of these threats, I ask Joyce about what role Google itself has to play in the wider security space going forward - is it being a first response system, a protector - or to take the first strike?

“That is the goal,” she says, “we do take direct action, especially if they’re touching the Google infrastructure - but we have a lot of options to take action…more and more, some of the creative thinking we have is, how do we disrupt this type of activity - within the laws that govern this type of activity.”

Working with law enforcement forces is a key method, she notes, but Google Cloud also takes direct action on the infrastructure itself, and partners with other organizations for co-ordinated takedowns.

“There's a lot of ways we can disrupt and do the right thing,” she says, highlighting the company’s responsibility to protect not only Google’s products and people - but its customers too, “the more we know about the threats, the more we can do.”

I also ask Joyce about the role of AI in cybersecurity, given it has transformed so many other areas in the business world over the past few years.

The company announced several AI-enabled security services and tools at Cloud Next 25, most notably Google Unified Security (GUS), a combined platform for firms to access all their security tools in a single location, as well as several security-focused AI agents.

Joyce says the potential impact is, “fascinating…this is now the modern way people are going to expect to be able to interact with data.”

She notes that threat detection, analysis and mitigation will all see a huge boost from AI, greatly speeding up processes that used to take months into a matter of days, all enabled by natural language prompts that make it easy for all workers to use.

"I don’t think that we have an excuse to not lead in this space,” she adds, "because we have the technology, we have the expertise, we have the recipe to make something incredible.”

• AWS, Microsoft and Google testify at UK CMA investigation
• Microsoft’s licensing terms and fees appear to be a problem
• Google denies dominance, says it’s a challenger

Ever since the UK’s Competition & Markets Authority has been investigating the health of the UK cloud market, the hyperscalers in question have been battling it out against each other by submitting little digs in official documentation meant to act in their defense.

This time, AWS is slamming Microsoft for its unfavorable licensing terms, which are preventing customers from being able to use alternative cloud storage providers without excessive fees.

“Perhaps 50% of those workloads currently running on Azure would move elsewhere if it was economically feasible,” AWS told the CMA.

AWS criticizes Microsoft’s licensing terms and high fees

Microsoft’s 2019 changes increased the costs relating to running Windows Server on non-Azure cloud (such as AWS, Google Cloud and Alibaba Cloud), making it up to four times more expensive for customers.

Those fees aren’t just affecting customer, either. “AWS said that it has to offset the additional costs imposed by Microsoft’s licensing restrictions including on the cost of licences that need to be repurchased and the additional monetary impact of non-pricing features,” the CMA hearing summary continues, but “it cannot offset all those costs in a profitable way.”

Google also agreed with AWS’ comments in a separate hearing summary, providing an example where a customer chose Azure solely for licensing/commercial reasons despite having an actual preference for Google Cloud.

However, Google might not be entirely on Amazon’s side. The company described itself as a “challenger cloud provider and a distant third in size to the two market leaders, AWS and Microsoft.” Still, Google is worried that Microsoft could dominate the UK cloud market in as little as five years “if nothing change[s].”

Defending its position, Microsoft argued that egress fees are not a major concern for customers, adding that even after removing the fees under the EU Data Act, switching remained low.

The outcome of the investigation is anyone’s guess, but the market dominance is clear.

When Ofcom first submitted its complaint to the CMA in October 2023, it noted in a report that Microsoft and Amazon accounted for 70-80% of the UK’s cloud market in 2022, estimated to be worth £7 to £7.5 billion. Google, in third place, only accounted for 5-10%.

You might also like

• We’ve listed the best cloud hosting providers
• UK government says Microsoft and AWS unfairly dominate the cloud market
• Keep your data online with the best cloud storage and best cloud backup tools

• Google Cloud’s us-east5-c (Columbus, Ohio) zone had a six-hour outage
• It was caused by a failure to its uninterruptible power supply
• Over 20 services and some storage disks went down

Much like the unsinkable Titanic met its demise when it sank, Google Cloud has recovered from a major outage caused by an interruption to its uninterruptible power supplies (UPS).

The company confirmed that its us-east5-c zone, otherwise known as Columbus, Ohio, experienced “degraded service or unavailability” for a period of six hours and 10 minutes on March 29, 2025, blaming it on a “loss of utility power in the affected zone.”

Over 20 cloud services suffered reduced performance or downtime as a result of the outage, including BigQuery, Cloud SQL, Cloud VPN and Virtual Private Cloud.

Google’s uninterruptible power supply just had a pretty major failure

In its incident report, the company explained exactly what had happened: “This power outage triggered a cascading failure within the uninterruptible power supply (UPS) system responsible for maintaining power to the zone during such events.”

“The UPS system, which relies on batteries to bridge the gap between utility power loss and generator power activation, experienced a critical battery failure,” the log continues.

Google’s Columbus zone uses powerful chips from Intel like Broadwell, Haswell, Skylake, Cascade Lake, Sapphire Rapids and Emerald Rapids, as well as the AMD EPYC Rome and Milan processors to power its cloud computing services. The cloud giant also noted that “a limited number of storage disks within the zone became unavailable during the outage.”

Engineers were made aware of the outage at 12:54 PT on March 29, successfully bypassing the failed UPS and restored power via generator by 14:49 PT. Most services were brought back online pretty quickly, thereafter, but some manual action was required for a full restoration hence the six-hour outage.

Google now promises to learn from this event, hardening cluster power failure and recovery paths and auditing systems that did not automatically failover, as well as working with its UPS vendor to mitigate future incidents.

You might also like

• Check out the best cloud computing services
• Oracle quietly confirms public cloud data breach, customer data stolen
• Keep your data online with the best cloud storage and best cloud backup tools

With AI agents becoming an increasingly common sight in businesses everywhere, Google Cloud has become the latest major company to ramp up its efforts in the space.

At its Google Cloud Next 25 event, the company unveiled several upgrades to its Agentspace platform to make agent discovery and adoption easier.

Just to give things an extra boost, Google Cloud also announced a new partnership with Nvidia designed at making its offerings even more intuitive.

Google Agentspace expansion

Following the initial launch of Google Agentspace in December 2024, the updates were mainly aimed at making creating and deploying AI agents easier

This includes giving employees access to Agentspace search and analysis tools directly from the search box in Google Chrome. The multimodal search capabilities can help track down exactly the data needed within your business, or give customers access to the answers they need.

The search results can cover content from the web, or from your business' most commonly-used apps and software, including the likes of Google Workspace, Microsoft 365, and apps like Jira, Salesforce, or ServiceNow.

Elsewhere, workers can also use a new Agent Gallery to find and deploy new agents quickly as well as creating their own agents with the new no-code Agent Designer platform - or launch some of Google's latest own-brand offerings, Idea Generation agent and Deep Research agent.

In order to make sure all these new agents co-exist effectively, Google Cloud has also launched a new interoperability protocol called Agent2Agent, which it says, "will allow AI agents to communicate with each other, securely exchange information, and coordinate actions on top of various enterprise platforms or applications."

Built on existing standards to allow easier integration, the company has already signed up more than 50 partners for the launch, including enterprise heavyweights such as Salesforce, PayPal, Box, Atlassian and more.

Finally, to harness the power of some of the most powerful computing hardware around today, Google Cloud and Nvidia have signed a collaboration bring the former's AI models to Nvidia Blackwell HGX and DGX platforms, as well as Nvidia Confidential Computing.

“By bringing our Gemini models on premises with Nvidia Blackwell’s breakthrough performance and confidential computing capabilities, we’re enabling enterprises to unlock the full potential of agentic AI,” said Sachin Gupta, vice president and general manager of infrastructure and solutions at Google Cloud.

“This collaboration helps ensure customers can innovate securely without compromising on performance or operational ease.”

You might also like

• Developers are pretty optimistic about AI agents
• AI agents are proving remarkably popular - but firms still face a host of challenges
• These are the best collaboration tools for businesses

Google Cloud CEO Thomas Kurian has said the company looks poised to go from strength to strength as the global demand for more AI technology continues unabated.

Fresh from hosting the opening keynote at the company's Google Cloud Next 25 in Las Vegas, Kurian told a media Q&A session attended by TechRadar Pro about the latest challenges facing the business, as well as its myriad of successes.

"(AI) is simplifying technology itself," Kurian said in the session, "You have the same technology that runs Google, available to every small business...by simplifying the technology, you bring it to everybody."

"Extremely dynamic" tariff situation

"If you look at the highest level of what we're trying to do - it's relatively simple," Kurian noted, highlighting how Google Cloud has expanded its global footprint to cover 42 regions across the world, with more to come soon.

"Our success is reflected not by what we do, but what customers and partners do with our technology," he added.

"We are very well-positioned due to the breadth of our portfolio - but also because of the differentiation in what we're doing with AI."

Asked about the possible effects of US tariffs on business around the world, Kurian replied with a smile that the tariff discussion is "an extremely dynamic one", and that Google Cloud has been through many cycles like this (most recently in the Covid crisis) and that he was confident the company will be able to navigate this period too.

"There's many, many things we do as a technology provider - we are working as a global company, within a geopolitical context, and our solution to that is to work with governments to adapt our technology to government regulations, so that customers can use it in different markets."

Kurian was also asked about how Google Cloud plans to deal with the rising energy costs caused by the growth in demand for AI.

Stating the company has done "a lot of work" over the past two years to "hugely reduce" the cost of training and inferencing of models, he noted the company will continue to optimize the cost of serving models while improving quality.

Kurian added Google Cloud has more than seven times the water-cooled AI systems as the rest of the world combined, and was also working on implementing nuclear power and sustainable energy sources such as hydroelectric and solar.

"For us, it's really important that people see AI as being a technology that can also drive both efficiency in the consumption of energy, and also created new forms of energy because of the investments that we are making."

You might also like

• "The opportunity with AI is as big as it gets" - Google CEO Sundar Pichai pledges $75 billion cloud, AI spending spree
• Alphabet CEO says more than a quarter of new code at Google is AI-generated
• Check out the best AI writers around today

Our live coverage of Google Cloud Next 25 has now ended!

We were live at the Mandalay Bay resort in Las Vegas for a conference packed with Google's latest cloud and AI news.

If you missed any of the action, don't fear - you can still catch up with everything we saw below...

Good morning from sunny Las Vegas! We're here for Google Cloud Next, and after a slighted delayed arrival into the city last night, are busy getting ready for the event kicking off tomorrow.

If you'd like a reminder of everything announced at last year's event, you can revisit our live blog here.

Before the conference kicks off tomorrow, we've been invited for a VIP press conference at an undisclosed location in Las Vegas...

Will we be allowed to report on it? What will it be about? Is this just a big AI prank? Stay tuned...

Well, that was interesting - we were whisked off to the famous Las Vegas Sphere, for a look at how Google Cloud worked to help prepare the movie classic The Wizard of Oz for the venue, launching on August 28.

It's fair to say it will be a spectacular show, with the original movie expanded, using AI content generation, to spread across the 16K LED screen at the Sphere.

The event was introduced by Alphabet and Google CEO Sundar Pichai, who welcomed us to the show, and outlined the scale of the technology being used.

Make no mistake - seeing a 75-year-old film, originally filmed in a rectangular format, expanded and spread across the screen, is pretty spectacular...

The event then wrapped up with a discussion between Google Cloud CEO Thomas Kurian and Jim Dolan, CEO of Sphere Entertainment Co, covering how the partnership came about, and their hopes for AI in the future.

Kurian then took to the stage solo to welcome us to Google Cloud Next 25.

That should be that for today - but tune in tomorrow for the first official day of the conference, with press Q&As and keynotes to come!

Good morning and welcome to the first day of Google Cloud Next!

After last night's magical Sphere experience, we're not sure how Google could possibly top that - but if anyone can, they can - and we are in Vegas after all...

We're off for breakfast and caffeine, before the keynote starts at 9am PST.

We're en route to the keynote now - as you might expect, it's pretty busy...

We're in and seated! As per Vegas conference rules, we're being blasted with upbeat techno - accompanied with strange and fantastic animated scenes generated by Google's Veo2 model...

Brass instruments floating over the Las Vegas strip and the Grand Canyon? It's amazing what AI can do...

Here we go - the lights go down and it's time for an intro video, again showing some...interesting visuals generated by Veo 2.

Google Cloud CEO Thomas Kurian takes to the stage.

"The future is being built by all of us," he notes, saying Google Cloud has shipped over 3,000 improvements and updates in the past year alone.

"Google's AI momentum is exciting," he adds, outlining huge adavnces in adoption of Gemini, Imagen, Veo and the company's other AI tools.

Our first special guest - Sundar Pichai, CEO of parent company Alphabet.

He talks us through the Sphere/Wizard of Oz collaboration we saw last night, as an example of what technology can do to show what is possible.

"With Google Cloud, we see AI as the most important way we can help advance your mission - the opportunity with AI is as big as it gets," he notes.

Pichai adds that Alphabet is ramping up investment in AI infrastructure - and reveals the first news of the day - Google is making its Cloud Wide Area Network open to businesses across the globe.

That's not all - Pichai unveils Ironwood, the company's next-generation TPU.

The seventh-gen hardware will help power the next generation of AI computing, allowing for more analysis and inference like never before.

Moving quickly on - Pichai outlines the success of Gemini 2.5 Pro, which he says is streaking ahead of the competition.

He shows off a couple of examples of why Gemini 2.5 is a "significant leap" over its rivals, including simualting a Rubiks cube - but also a massive simulation of the Earth's magentic field, showing off the huge potential of the model.

Pichai reveals Gemini 2.5 Pro is now available for everyone in AI Studio, Vertex AI, and the Gemini app.

There's also the release of Gemini 2.5 Flash, offering even more access to the platform.

Pichai notes Gemini is helping power a wide range of tools, such as NotebookLM, used by over a million businesses.

There's also a mention for Veo 2, which is being used by advertising and marketing agencies across the world, Pichai says.

We're then treated to a video showing how McDonald's is using Google's AI tools, using real-time data to improve performance and boost the overall restaurant experience - tasty stuff.

Kurian returns to the stage, highlighting why customers want to work with Google Cloud.

The company's AI-optimized platform is a major selling point, he adds, with the ability to build new agents that offer more help than ever before.

We welcome Amin Vahdat, VP and GM, ML, Systems and Cloud AI, to the stage, to talk about hardware.

He's joined on stage by an actual Ironwood TPU, noting how it can deal with even the most demanding AI workloads.

Ironwood supports up to 42.5 Exaflops, well beyond even the biggest supercomputers around today.

This will be backed up by Nvidia GB200 Blackwell chip-powered hardware - and Google will also offer Vera Rubin GPUS on Google Cloud.

There's also a new cloud and rapid storage options for greater throughput and latency.

Vahdat also runs through a host of new software announcements, including boosts for GKE Inference Gateway, Pathways on Google Cloud, and vLLM to TPU, giving easy TPU inference on a familiar library.

"We're truly seeing tremendous momentum," Vahdat notes, with Google helping power AI unicorns across the globe, as well as major customer usage across a wide variety of industries.

Google Distributed Cloud also gets an upgrade, with new Gemini tools allowing Google Distributed Cloud to run locally in air-gapped and connected envrionments.

A quick video from Nvidia CEO Jensen Huang is up next, highlighting the partnership between the two.

"No company is better at every sungle layer of computing than Google and Google Cloud," he notes.

Vahdat moves on to Google Workspace, which is getting a major boost from Gemini, which is now included in all subscriptions.

A new "Help me analyze" tool is coming to Google Sheets, along with "Audio overviews in Docs", and Google Workspace Flows, to help automate work with agents in the loop.

Imagen 3 is next up, now offering greater image generation than ever, along with voice-generation tool Chirp 3.

Vahdat announced Lyria is now available on Google Cloud, which can create music and other audio from just 10 second of input.

Veo 2 is also much more powerful than ever, offering first and last shot control, more camera angles, extra inpainting and outpainting tools, and much more.

All together, Vahdat says Google is, "the only company that offers generated media models across all modalities".

After that bombardment of information, it's time for a demo of just what Vertex AI can do.

To create a teaser video for the Google Cloud Next 25 closing party, Veo 2 is used to create hyper-realistic video input, including different angles, approaches and even audience members.

Kurian returns, and it's time to talk about Vertex AI.

AI models are being used in the physical world too, he notes, highlighting work done in robotics by Samsung and DeepMind.

Vertex allows users to discover, customize and deploy the best models for your business, he notes, introducing a customer example from Intuit, who used AI to help even the most complex tax filings a breeze.

Tens of thousands of companies are using Vertex to find the best models, Kurian notes - with use cases from patient healthcare to energy management.

Agents are never far away, with the tools now able to be easily built on Vertex, connecting to a number of top third-party partners.

Next up, Kurian wants to talk more about agents.

The company is already working with the likes of Salesforce, whose CEO Marc Benioff tells us via video how its Agentforce platform is a key ally for Google Cloud.

Google Cloud will now let users build and manage multi-agent systems, Kurian reveals.

The company's new Agentspace platform will let users scale adoption and accelerate deployment, wth a new Agent Development Kit simplifying the building of multi-agent systems - powered by Gemini, of course.

There's also a new Agent2Agent protocol to let agents talk to each other, regardless of what model they were built on.

We're now seeing how Agentspace works in action, with a demo showing off how agents can be used to automate everyday tasks.

Agentspace is now integrated into the Chrome browser, Kurian reveals, meaning you can search company data directly within your browser.

We're going even more in=depth on agents now, taking a look at several key areas where the technology in making a difference.

From customer agents built with Vertex AI Search used by Reddit Answers, to medical apps finding patient histories, to retail companies delivering personalized recommendations, the potential seems massive.

Next are some new offerings to improve Agents via Customer Engagement Suite, including streaming video support, new voices, and AI assistance to build agents.

We then see a demo of these new tools in action, helping get a customer the items they need to complete a garden, providing custom recommendations via live video in real time.

Industry agents are also a key use case, no matter how specific.

Wendys, Mercedes-Benz and The Home Depot are among the companies using Google Cloud technology to make their agents more helpful and intuitive.

Creative Agents are another key area, augmenting creative teams to allow content production at scale, or reimagining existing content.

We're once again looking at the Wizard of Oz Sphere experience, which used Google AI, including Veo 2, to turn the 1939 film into a fully-immsersive entertainment experience in super-fast time.

Agencies across the world are using Google's models to help create ads, improve localization, and reduce production times across the board.

Google seems very keen to ensure that models are working alongside human workers to be creative, rather than completely replace...

Data agents are also set to be a major boost for customers, taking lots of the heavy lifting and time-intensive tasks away from busy workers.

The likes of Mattel are working with Google Cloud to bring together customer feedback to improve data collection, collecting sentiment in real time to identify trends quicker than ever.

Google's BigQuery platform offers more reach than any competitor, the company says, bringing together structured and unstructured data to build the most useful agents for your business.

Next up, we're looking at Code Agents - as Gemini can be a highly effective partner using its Code Assist tools.

An enterprise-focused version is already proving a major success, helping companies around the globe.

Today new agents are being launched for Google Code Assist, including greater insights on the task being worked on, and more interactivity options.

Now, we're moving on to security, with Sandra Joyce, VP of Threat Intelligence.

She's straight in to Security Agents, with new critical cyberdefenses on offer for users now thanks to Google Threat Intelligence, Google Security Operations, Cloud Security Command Center, and Mandiant services.

New Malware Analysis and Alert Triage security agents are being announced today, giving customers even more tools to spot threats before they do serious damage.

A big one - Google Unified Security is a new converged security solution bringing together all of the company's tools and features.

We're treated to a demo of the new platform - named GUS for short - can track all of a company's workflows, combined with the latest threat intelligence, to triage the incident, create an automated response and shut down affected systems - all in a much shorter time than previously possible.

GUS looks set to revolutionize how companies ensure they stay protected, offering real-time advice and recommendations defined to even the smallest threats...

We're nearly at the end, and Kurian returns to the stage to wrap things up.

"We're delivering an amazing stream of new innovations, and making it easy to integrate those innovations into your existing technology landscapes," he concludes.

"What an amazing time for all of us to experience and work with these technology advances."

"We are honoured to be building this new way to cloud, with all of you."

And that's a wrap! We're off to digest all of the mountain of new announcements made in that keynote, so stay tuned to TechRadar Pro for write-ups of the biggest announcements coming soon.

We've got our heads around some of the news from this morning, so if you missed it - here's some of the biggest announcements...

Google Unified Security brings the power of AI to your security suite

The company has unveiled Google Unified Security, a new unified security platform designed to take the stress out of keeping your business safe from the latest threats.

Google Unified Security, affectionately known as GUS, promises a major step forward in threat detection and mitigation, and is outfitted with the company's latest Threat Intelligence, AI tools and services to keep users safe.

Google unveils new security AI agents to keep your business safe from the latest threats

Part of the new Google Unified Security platform, the company revealed two new AI security agents.

First is an Alert Triage agent in Google Security Operations which is able to carry out dynamic investigations on behalf of users, reducing the workload of analysts who otherwise are triaging and investigating hundreds of alerts per day.

The other new addition is a Malware Analysis agent, part of Google Threat Intelligence, which can investigate whether code is safe or harmful. It will also have the ability to create and execute scripts for deobfuscation, before summarizing its work and offering up a final verdict.

Google Cloud unveils Ironwood, its 7th Gen TPU to help boost AI performance and inference

Ironwood is the company's 7th-generation Tensor Processing Unit (TPU).

At its top-end Ironwood can scale up to 9,216 chips per pod, for a total of 42.5 exaflops - more than 24x the compute power of El Capitan, the world's current largest supercomputer.

Each individual chip offers peak compute of 4,614 TFLOPs, what the company says is a huge leap forward in capacity and capability.

We also heard from Alphabet and Google CEO Sundar Pichai at the keynote, who reaffirmed the company's plans to spend big on infrastructure and technology going forward.

"The opportunity with AI is as big as it gets" - Google CEO Sundar Pichai pledges $75 billion cloud, AI spending spree

And that's a wrap on our live coverage from day one of Google Cloud Next 25 - it's been a busy day, so we're off to rest and recuperate, before heading back tomorrow!

Be sure to join us then, as we'll have plenty more exciting content and news from the show...

Good morning and welcome to day two of Google Cloud Next 25!

Yesterday was packed full of news and announcements, and we're expecting more of the same today.

First up, we have a media Q&A with Thomas Kurian and others, before the developer keynote later today, so stay tuned for all the news from those events and more!

First up today, we have a roundtable with some pretty big names, including Google Cloud CEO Thomas Kurian, alongside Tara Brady, President, Google Cloud EMEA, and Eduardo Lopez, Vice President, Google Cloud LATAM.

The panel is here, and we have some introductory comments from Kurian on what's been the most important announcements from the show for him.

He notes how Google Cloud has now expanded to 42 countries across the globe, and runs us through the biggest news from yesterday.

"Our success is reflected not by what we do, but what our customers do with our technology," Kurian notes, shouting out some customer success stories from the EMEA and LATAM regions.

"We are very well-positioned."

Brady notes the growth has been "phenomenal" in EMEA over the past year.

The first question is about how Google Cloud deals with complex geo-political situations, and if this affects its business outlook.

Kurian notes governments around the world want more control over sovereign technology, and allows controls to me laid over the top of what Google Cloud makes available.

He adds its Sovereign AI systems are controlled in a local environment, managed by a national foundation - and that the company works hard on localizing its products for different countries.

Asked about the EMEA region, Brady remarks that security is a crucial consideration, along with flexibility - and of course AI, and the competitive advantage this offers.

Next is a question about how companies like Google Cloud can cope with the energy demands caused by AI expansion.

Kurian says the company has done a lot of work on reducing the energy usage on its systems - with the costs and demands dropping fast.

"You'll see us continue to optimize the cost of serving models, while improving quality," he notes.

A pretty staggering stat from Kurian - Google Cloud has seven times the amount of water-cooling systems than the entire rest of the world, combined.

Wind, nuclear, hydro-electric and solar are all set to play a key role in the future, he adds.

"For us, it's really important people see AI as a key technology in driving the consumption of clean energy," he says.

Unsurprisingly, a question about the US tariffs is next - how will Google Cloud look to deal with the possible effects?

Kurian replies with a smile that the tariff disucssion is "an extremely dynamic one", and that the company has been through many cycles like this - most recently in the Covid crisis - and he's confident the company will be able to navigate this period too.

Kurian is asked about UAI tools - he replies the entire approach is about simplifying technology for users everywhere.

Bringing AI is the next step - "it is simplifying technology itself," he notes.

"You have the same technology that runs Google, available to every small business...by simplifying the technology, you bring it to everybody."

Next is a question about Agentspace, and how it will affect companies which already have a presence with other ecosystems.

Kurian notes Agentspace came from the company's own observations that workers were spending so much time just looking for information.

The next question is about building new data centers - how does Google Cloud decide where its next projects will be?

Kurian says it is a massively complex decision - but key criteria includes customer demand, availability of power (especially renewable), and space.

"You'll see us continue to expand our global footprint," he says.

Another question on tariffs - and if this will lead to national clouds in the future, especially if the US (or other countries) suddenly decide to cut off access?

Kurian notes customers can already benefit from localization functions, with data kept seperate even from Google itself. The company also offers services for those customers worried about privacy or access to data, and offers sovereign partnerships with European organizations, which can supervise on top of Google's operations.

"The proof is in how many customers are choosing to run (Google Cloud)".

The final question is about Google's proposed acquisition of Wiz - is it really worth that much?

"Wiz solved a problem that was really important for customers", Kurian notes, noting Wiz had been a long-term partner for some time.

And that's a wrap on the Q&A - we're off to get some coffee, but we'll be back for the developer keynote later today.

We're back, and headed to the day two keynotes - this one is for the developers, so expect a deeper dive into some of the major announcements we saw yesterday - and possibly some extra surprises, too...

We're nearly there - but going to be another packed turn-out, it looks like...

We're in and seated - and it's time to kick off!

The lights go down, and Brad Calder, VP & GM, Google Cloud, takes to the stage.

"This is what we've been waiting for," he says, outlining's Google Cloud's advancements in three main areas.

Its new Agent Development Kit, Agent Engine and Agentspace all get an early shout-out.

"Together, these allow you to create agents where AI and humans can work together to reach a common goal."

All of this, of course - is powered by Gemini.

Calder welcomes Richard Seroter, Chief Evangelist, Google Cloud, and Stephanie Wong, Head of Developer Skills & Community, Google Cloud, to the stage to take a closer look.

From getting started, to scaling, to a sneak peek at the future, we're about to take a look at a lot of demos...

Our first demo looks at how you can build AI-enabled apps with Gemini in AI Studio.

Namely, how Gemini can help our friendly demo team remodel their kitchen - including huge amounts of detail, creating a custom prompt despite being large amounts of unstructured data.

As the model thinks, we're given a run-through of exactly what is going on behind the scenes - including the in-depth "Thinking" box.

The end result is a beautifully-designed new kitchen idea, which can then be customized even further, and take in the latest Google search data for real-time details on information such as building material costs.

Gemini's long-context windows are getting a shout-out, giving users much more ability and flexibility to create incredibly detailed prompts, and outputs.

Gemini models in AI Studio offer not just advanced multimodal reasoning, but also mixed modalities for inputs and outputs.

Next, we're moving on to how agents can help businesses themselves.

Agents can go from the smallest to the largest tasks, offering unique capabilities when it comes to customization.

Our next demo is of Agent Development Kit - we're going to see just how easy it is to build an agent.

To build an agent, we need three things - an instruction, the tools, and the model.

The demo shows how easy it is to connect to different models, using natural language instructions to carry out its task.

With all that done, a final document is created in seconds - a huge time-saver for users everywhere.

Next up, we're looking at how developers are building with Vertex AI Agent Engine, now generally available.

This allows users to run fully managed enterprise-grade agents, along with advanced security.

We're then on to Google Agentspace, the new management platform for agents, allowing quick creation, editing and deployment of agents, along with third-party connectors - and then we're on to building multi agent sytems in Vertex AI.

We're being shown a demo of how easy it is to build even complex, multi-source systems in ADK.

Users can also deploy directly from ADK to Vertex AI Studio, with deployment taking just a few minutes.

The demo then shows how straightforward debugging a troublesome agent can be - with just a few steps to get the tool up and running again.

A new cloud investigations tool can help determine where the errors in your code or infrastructure might be, before showing the users in an easy-to-understand sheet, including suggested fixes.

Next, we're moving on to how Gemini can be useful in IDEs and tools for developers.

We're taking a look at how different IDEs can benefit from Gemini to create code powered by AI.

The company is allowing devs to use Gemini wherever they create code, and we're then taken to a look at how different models for different use cases can be accessed.

Vertex's AI Model Garden shows off all the latest and greatest offerings out there, allowing users the chance to try out different options to see what the final result looks like.

Now, we're taking a look at a customer use case, between Google Cloud and Major League Baseball.

The companies have worked on a huge data-sharing project, and we're introduced to the winner of a Google hackathon to design the next greatest MLB tool.

Our winner used Gemini and Google Cloudto design a video analysis tool, without the need for high-speed cameras, to look at pitcher performance over time.

We're now moving towards what's next for Google Cloud...first up, the next generation (already) of data agents.

Turning raw data into a data app is soon going to be as easy as a few prompts, with BigQuery bringing together various data sources for quick and easy analysis.

Our next shift is to look at the future of software development through agents.

We're shown a demo of how agents powered by Gemini Code Assist can help build the next generation of apps - moving beyond editors to a new way to develop software, with agents.

Code Assist is getting way more customization options and capabilities, giving developers all kinds of ways to tailor apps exactly how they want them.

And that's a wrap on the day two keynote - and for Google Cloud Next 25 as a whole.

Stay tuned for more of our wrap-up coverage, but thanks for joining us over the last few days, and thanks for reading TechRadar Pro!

• Google will acquire Wiz in a $32 billion deal
• The pair say move will boost cloud offerings and lower customer costs
• Google will be able to better compete with Amazon and Microsoft

Google has announced it will be acquiring Wiz in a $32 billion deal that will help boost its cloud computing offerings.

“This acquisition represents an investment by Google Cloud to accelerate two large and growing trends in the AI era: improved cloud security and the ability to use multiple clouds (multicloud),” Google said in a statement.

Google has an extensive cloud infrastructure portfolio, and Wiz already offers cloud security on rival platforms such as Amazon Web Services, Microsoft Azure, and Oracle Cloud, so the move will help boost security and lower costs for organizations of all sizes, Google says.

Strengthening the cloud

Businesses are increasingly turning to cloud services due to the convenience of not needing on-prem hardware to run resource-hungry tools and applications, with the AI boom being a big contributing factor. The acquisition will allow Google to better compete with Amazon and Microsoft in the cloud enterprise environment.

Google previously offered $23 billion for the platform in 2024, but negotiations fell apart over discussions about how much independence Wiz would have in its development direction under Google, and the potential for regulatory and antitrust roadblocks.

However, President Trump’s recent rhetoric around removing some of the red tape attached to big tech acquisitions may be a cause for optimism that this new deal will go through, despite the potential for regulatory challenges.

“Google Cloud and Wiz share a joint vision to make cybersecurity more accessible and simpler to use for organizations of any size and industry," Thomas Kurian, CEO, Google Cloud said.

"Enabling more companies to prevent cyber attacks, including in very complex business software environments, will help organizations minimize the cost, disruption and hassle caused by cybersecurity incidents.”

Assaf Rappaport, Co-Founder & CEO, Wiz also commented on the announcement, "Wiz and Google Cloud are fully committed to continue supporting and protecting customers across all major clouds, helping keep them safe and secure wherever they operate. This is an exciting moment for our company, but an even more important one for customers and partners, as this acquisition will bolster our mission to improve security and prevent breaches by providing additional resources and deep AI expertise.”

You might also like

• Take a look at my picks for the best cloud storage
• I've rounded up all the best cloud backup solutions
• Salesforce and Google Cloud sign multi-billion dollar software deal which could affect how you use AI at work

• Salesforce has ramped up its AI model sharing with Google
• Partnership comes as Salesforce bets big on agentic AI
• Long context windows and real-time data are just some of the benefits

Salesforce and Google Cloud have announced what they describe as a “major expansion” of their strategic partnership – a two-way deal which will see them sharing cloud and AI resources.

Central to the deal is how Salesforce customers will nowbe able to build AI agents using Google’s Gemini models as well as deploy Salesforce on Google Cloud.

The news comes during the emergence of agentic AI, a sector that’s estimated to be worth $2 trillion.

“Our mutual customers have asked us to be able to work more seamlessly across Salesforce and Google Cloud, and this expanded partnership will help them accelerate their AI transformations with agentic AI, state-of-the-art AI models, data analytics, and more," noted Google Cloud CEO Thomas Kurian.

Salesforce and Google Cloud deepen partnership

The news comes during the emergence of agentic AI, a sector that’s estimated to be worth $2 trillion.

According to Salesforce data, 84% of CIOs believe artificial intelligence will be as significant as the internet itself.

Salesforce’s access to Google’s models includes Grounding with Google Search through Vertex AI, which means that users will be able to access “up-to-the-minute data, news, current events, and credible citations.”

Salesforce also noted the importance of giving its customers the freedom to choose their own models – this flexibility should in theory prevent them from seeking alternative solutions, keeping them spending their money with Salesforce.

“Through our expanded partnership with Google Cloud and deep integrations at the platform, application, and infrastructure layer, we’re giving customers choice in the applications and models they want to use," noted Salesforce Chief Engineering Officer Srini Tallapragada.

The announcement also references Gemini’s two million-token context window, making it one of the most powerful models for contextual understanding and reasoning.

The two companies have promised deeper integrations across Data Cloud, BigQuery, and Cortex Framework later in 2025, as well as new native Tableau, Looker, and BigQuery integration.

You might also like

• Salesforce unveils its vision of AI agents to help every business
• Check out the best AI tools and best AI writers you can use today
• We’ve listed the best cloud computing services for accessing powerful chips

• Google Cloud now offers PQC algorithms for digital signatures
• The move comes as part of Google Cloud quantum-safe roadmap
• The feature is available in preview now, with more PQC changes to come

Google Cloud has taken an additional step to secure its platform from the threat posed by quantum computing to public-key cryptography by announcing the availability of quantum-safe digital signatures in preview in the Google Cloud Key Management Service (KMS).

Theoretically, quantum computers will be able to crack most modern encryption algorithms in a matter of hours, but the viability of such a system is still at least a decade away by most expert estimates.

This does present threats today, however, with cybercriminals stealing vulnerable encrypted data today in order to decrypt it when ‘Q-Day’, or the day when the availability of cryptographically-relevant quantum computers arrives. This has necessitated the development of post-quantum cryptography (PQC) to secure data against ‘harvest-now-decrypt-later’ attacks.

Securing the future of digital signatures

Quantum-safe digital signatures are just one step in Google Cloud’s roadmap towards a fully PQC platform.

For digital signatures specifically, quantum brings the threat of forgery and tampering, with Google Cloud highlighting that working to resolve these threats now will help build a more resilient infrastructure and inspire other platforms to do the same.

Google Cloud has also revealed in the announcement that to help enforce transparency and code-auditability, the quantum-related software implementations will be available as part of Google’s open-source cryptographic libraries BoringCrypto and Tink.

This update to digital signatures specifically implements the ML-DSA-65 and SLH-DSA-SHA2-128S PQC algorithms in line with NIST standards. Google Cloud will also be implementing new PQC algorithms and techniques into its Hardware Security Modules (HSM).

Google Cloud customers will be able to use the latest PQC algorithms in Cloud KMS and provide feedback on their implementation, and suggestions on other specific cryptographic requirements. Essentially, the use of Cloud KMS’s latest algorithms acts as a testbed for building and expanding the availability of PQC across the Google Cloud platform.

You might also like

• These are the best encrypted messaging app
• Take a look at our guide to the best password manager
• Microsoft set to rival Google, IBM with first-of-its-kind quantum chip with topological core architecture

• The GE Profile Smart Grind & Brew is a drip coffee maker with an AI assistant
• The assistant can adjust brewing settings based on the questions you ask
• It's available to buy now from Amazon for $379 (about £300 / AU$600)

Just how smart can a filter coffee machine really be? Very smart indeed, it turns out. GE Profile (the premium division of GE Appliances) has just launched the Smart Grind & Brew – a drip machine that uses Google Cloud AI to help you brew the perfect cup each morning.

The GE Profile Smart Grind & Brew is a slim machine with a built-in conical burr grinder, which means you can enjoy freshly-ground coffee without investing in (and finding space for) two appliances. It also minimizes potential mess because the ground coffee falls directly into the brew basket at the correct dosage, with no weighing or transferring.

The machine also has a built-in water filter, so you can use water directly from the tap without it affecting the flavor of your drink, and helping you get a more consistent taste.

Another particularly nice feature is the inclusion of a removable brewing stand (the part that holds the jug or cup) so the machine can dispense coffee directly into a travel mug. Combined with the built-in timer, this would make it easy to have a fresh filter coffee to take to work, minimizing the risk of caving to temptation and buying a takeaway coffee en route.

What really makes the Smart Grind & Brew different to the best drip coffee makers we've tested, however, is the SmartHQ AI Coffee Assistant, which is due to roll out to the machine with a software update very soon.

This lets you control the machine remotely using the SmartHQ smart home app, and have a "conversational AI experience" with the assistant. It also enables you to create custom profiles, receive real-time brewing feedback, and "is equipped with intuitive learning that automatically adjusts brewing based on questions asked."

SmartHQ will also let you add your coffee maker to your Amazon Alexa or Google Home smart hub, so you can operate it with voice controls.

So far, so smart

It all sounds very impressive, and the GE Profile Smart Grind & Brew is available to buy now from Amazon for $379 (about £300 / AU$600), which seems to be typical for a connected drip coffee machine.

Is there a catch? Well, although it has an insulated carafe, GE Profile doesn't mention the machine having a hotplate to keep your brewed coffee warm until you're ready for your next cup. That's the same problem I found with another high-tech drip machine I tested recently: the Fellow Aiden Precision Drip Coffee Maker. Although the Aiden could brew an excellent drink, its double-wall insulated carafe just didn't keep it warm long enough without a heat source underneath.

Hopefully that won't be quite so much of a problem with the Smart Grind & Brew, which (unlike the Aiden) is capable of dispensing coffee into other vessels, such as an insulated bottle, which will keep it at a drinkable temperature longer.

You might also like

• 5 hot coffee innovations from CES – from condensing espresso machines to super-fast cold brew
• The future of coffee - 3 hot trends brewing in 2025
• I took a barista class, and these are the most important lessons I learned for making perfect coffee at home

• Trillium has hit general availability just months after preview release
• Powerful AI chip offers more than four times the training performance
• Google uses it to train Gemini 2.0, the company's advanced AI model

Google has been developing Tensor Processing Units (TPUs), its custom AI accelerators, for over a decade, and a few months after being made available in preview, has announced that its sixth-generation TPU has reached general availability and is now available for rent.

Trillium doubles both the HBM capacity and the Interchip Interconnect bandwidth, and was was used to train Gemini 2.0, the tech giant’s flagship AI model.

Google reports it offers up to a 2.5x improvement in training performance per dollar compared to prior TPU generations, making it an appealing option for enterprises seeking efficient AI infrastructure.

Google Cloud’s AI Hypercomputer

Trillium delivers a range of other improvements over its predecessor, including more than four times the training performance. Energy efficiency has been increased by 67%, while peak compute performance per chip has risen by a factor of 4.7.

Trillium naturally improves inference performance as well. Google’s tests indicate over three times higher throughput for image generation models such as Stable Diffusion XL and nearly twice the throughput for large language models compared to earlier TPU generations.

The chip is also optimized for embedding-intensive models, with its third-generation SparseCore providing better performance for dynamic and data-dependent operations.

Trillium TPU also forms the foundation of Google Cloud’s AI Hypercomputer. This system features over 100,000 Trillium chips connected via a Jupiter network fabric delivering 13 Petabits/sec of bandwidth. It integrates optimized hardware, open software, and popular machine learning frameworks, including JAX, PyTorch, and TensorFlow.

With Trillium now generally available, Google Cloud customers have the opportunity to access the same hardware used to train Gemini 2.0, making high-performance AI infrastructure more accessible for a wide range of applications.

You might also like

• Google Cloud: No-one can deliver business AI value like us
• Google puts Nvidia on high alert as it showcases its rival AI chip
• Intel and Google Cloud team up to launch super-secure VMs