• Microsoft shareholders accuse the company of concealing AI spending risks
• Azure growth slowed as AI infrastructure demands consumed computing resources
• Investors claim key business challenges were not fully disclosed

Microsoft is facing a class action lawsuit from shareholders who say the company failed to disclose the financial fallout of its AI spending properly.

The complaint argues investors were given only part of the picture, omitting key details about spending requirements, cloud infrastructure limits, and the broader challenges tied to Microsoft's AI push.

The proposed class period stretches from May 1, 2025, to January 28, 2026, a stretch during which Microsoft shares hit record highs before sliding back down.

Shareholders challenge Microsoft's AI-related disclosures

The lawsuit, filed by The Rosen Law Firm on behalf of investors, claims Microsoft made misleading statements or simply left out information that mattered to its business operations.

According to the filing, company executives talked up Copilot's performance and the broader AI push while playing down concerns about cost and operational strain.

Court documents say Microsoft described Copilot as offering leading capabilities and strong adoption, language that helped keep investor confidence steady at the time.

The complaint goes further, alleging Microsoft never fully disclosed problems with user experience, interoperability, computational resources, internal organization, and data management.

Shareholders also argue that Microsoft's LLMs were falling behind certain competitors, requiring extra resources and development work just to keep pace.

According to the lawsuit, a meaningful part of computing capacity was pulled away from other revenue-generating services and redirected toward Copilot and AI research instead.

In fiscal year 2025, Azure revenue had grown 34% to more than $75 billion, and Microsoft kept telling investors that future growth would keep being driven by Azure.

Azure slowdown and rising expenditures draw investor scrutiny

Things came to a head after Microsoft's fiscal second-quarter 2026 earnings, covering the period through December 31.

The filing says Azure growth slowed unexpectedly and landed below what analysts had been expecting, raising fresh doubts about the company's infrastructure strategy.

During the earnings call, CFO Amy Hood reportedly attributed the slowdown largely to computational capacity constraints.

The complaint claims processor and graphics resources had been pulled toward Copilot and other AI models without much to show for it.

At the same time, Microsoft revealed capital expenditures of $37.5 billion for the quarter, pushing fiscal 2026 spending to $72.4 billion across just six months.

That figure already approached the $88.2 billion Microsoft spent across the entirety of fiscal 2025, according to the lawsuit.

Following those disclosures, shares fell more than $48 to $433.50, then kept sliding to $393 and eventually $380.

By the time the complaint was reported, Microsoft stock was trading around $399.76.

Microsoft executives, for their part, insist the company is doing everything it can to improve its AI tools and products.

Over the past year, Microsoft says it rolled out 625 new features, calling the product "way different than it was 90 days ago" as part of an effort to "improve the product rapidly."

Microsoft told Reuters the claims are "without merit" and said it "stands by the integrity of its public statements."

Via DataCenterDynamics

• 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.

• OpenAI explores an internal code platform after GitHub outages disrupt workflows
• GitHub’s Azure migration caused configuration problems and intermittent downtime for engineers
• Teams have been unable to commit code for hours during recent outages

OpenAI is reportedly developing an internal code repository platform after frequent outages on GitHub disrupted coding operations for its engineers.

The effort appears to be a direct response to GitHub's ongoing infrastructure migration to Microsoft Azure, which has introduced configuration problems and intermittent downtime.

Multiple outages over the past several months have left teams unable to commit code for hours at a time, raising concerns about reliability for AI-driven development workflows.

Rising outages and reliability concerns

A mid-year report from GitProtect indicated a 58% year-over-year increase in incidents during 2025, with 17 classified as major, resulting in over 100 hours of disruption.

The current migration has placed GitHub in a split-traffic state, with portions of its services still operating from its legacy Virginia data center, while other services have moved to Azure.

An early February 2026 four-hour outage was traced to an Azure problem, and a separate week-long outage was linked to a configuration change, which GitHub publicly acknowledged as “not yet meeting our expectations.”

These interruptions reportedly motivated OpenAI to explore a more controlled environment for its internal development and AI integration.

Employees familiar with the project said the platform could eventually be offered commercially, potentially bundled with Codex coding agents, although it may remain exclusive to OpenAI.

Developing internal code repositories is a common practice among large technology firms, with Google operating Piper and Meta using Sapling, but none have been commercialized.

OpenAI’s project, if made available to external customers, would represent a rare case of a company competing directly with a major investor.

Microsoft owns approximately 27% of OpenAI and acquired GitHub in 2018 for $7.5 billion.

OpenAI has already introduced ChatGPT features overlapping with Microsoft Office applications, suggesting broader strategic friction.

The move reflects broader concerns about the resilience of coding platforms under the growing strain of AI-driven workloads.

Massive AI data center buildouts are squeezing energy supplies, and the efficiency of internal collaboration tools is becoming more critical.

The planned platform may use advanced networking technologies such as Ultra Ethernet to improve data movement across AI infrastructure and reduce dependency on third-party platforms.

Engineers have reportedly experimented with laptops for programming tasks within this environment, integrating vibe coding tools to maintain workflow continuity despite external service interruptions.

OpenAI has not publicly confirmed the project, but the initiative reveals the tension between platform reliability and the growing demands of AI-powered development.

The platform remains in early development and could take months before any operational rollout.

Via Tom's Hardware

• Microsoft Azure Local can now run fully disconnected without needing the cloud
• European customers are increasingly demanding sovereign options
• CISPE welcomes the changes – looks to test them

Microsoft has announced three new updates to the company's sovereign offerings, including improvements across Azure Local, Microsoft 365 Local and Foundry Local.

The additions come as US-EU trade tensions continue, with Europe pushing for more digitally sovereign options.

European customers are increasingly demanding local infrastructure, protection from potential US CLOUD Act data access and reduced dependency on hyperscalers.

Microsoft offers three new sovereign options for the likes of Europe

Though Microsoft's Local additions don't respond to the desire to step away from hyperscalers, they do tackle data and infrastructure sovereignty demands while offering already-familiar experiences for existing customers.

In a blog post, Microsoft Specialized Clouds President and CTO Douglas Phillips explained organizations can now use on-prem infrastructure with Azure's governance and policy controls, all without cloud connectivity, in the form of the new Azure Local.

"Disconnected operations, management, policy and workload execution stay within the customer-operated environments," he wrote.

Previously, Azure Local needed periodic cloud connectivity. Staying offline for too long reduced functionality, but with the latest update, it can now run fully disconnected, making it ideal for highly regulated industries including defence and critical national infrastructure.

CISPE, Europe's body representing cloud infrastructure companies, welcomed the changes, promising to test Microsoft's changes against its upcoming Sovereign Cloud Services Framework to see if they pass.

Microsoft doesn't even stand out in the world of hyperscalers, with Google offering its own Google Cloud Airgapped solution and AWS selling its equivalent, the European Sovereign Cloud.

"Our approach helps customers to meet strict sovereign requirements in fully disconnected scenarios without compromising simplicity, while retaining flexibility where connectivity is possible," Phillips concluded.

• 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%.

• Nvidia integrates DeepSeek-R1 as a NIM microservice
• AWS supports DeepSeek-R1 with a focus on scalable and cost-efficient AI deployment
• Microsoft also has future local deployment plans for DeepSeek

Having taken the AI world by storm in recent weeks, DeepSeek has now made significant strides in expanding the accessibility of its advanced reasoning models.

The company has announced its flagship DeepSeek R1 model is now available on multiple platforms, including Nvidia, AWS, and GitHub.

DeepSeek’s open source nature allows developers to build models based on its architecture, and, at press time, there are 3,374 DeepSeek-based models available collaborative AI-model development platform Hugging Face.

Nvidia, AWS, Github & Azure now offers DeepSeek

On AWS, DeepSeek-R1 models are now accessible through Amazon Bedrock which simplifies API integration and Amazon SageMaker which enables advanced customization and training, supported by AWS Trainium and Inferentia for optimized cost efficiency.

AWS also offers DeepSeek-R1-Distill, a lighter version, through Amazon Bedrock Custom Model Import. This serverless deployment simplifies infrastructure management while maintaining scalability.

Nvidia has also integrated DeepSeek-R1 as a NIM microservice, leveraging its Hopper architecture and FP8 Transformer Engine acceleration to deliver real-time, high-quality responses.

The model, which features 671 billion parameters and a 128,000-token context length, utilizes test-time scaling for improved accuracy.

It also benefits from Nvidia’s Hopper architecture, using FP8 Transformer Engine acceleration and NVLink connectivity. Running on an HGX H200 system, DeepSeek-R1 can generate up to 3,872 tokens per second.

Microsoft’s Azure AI Foundry and GitHub have further expanded DeepSeek’s reach, offering developers a secure and scalable platform to integrate AI into their workflows.

Microsoft has also implemented extensive safety measures, including content filtering and automated assessments. The company claims that it plans to offer distilled versions of DeepSeek-R1 for local deployment on Copilot+ PCs in the future.

DeepSeek-R1 took the world by storm by offering a powerful, cost-efficient AI model with advanced reasoning capabilities and has dethroned popular AI models like ChatGPT.

R1 was reportedly trained for just $6 million, with its most advanced versions being about 95% cheaper to train than comparable models from NVIDIA and Microsoft.

You may also like

• Nvidia calls DeepSeek an 'excellent AI advancement' and praises the Chinese AI app's ingenuity
• OpenAI says DeepSeek used its models illegally, and it has evidence to prove it, new report claims
• DeepSeek and the race to surpass human intelligence

• BlackBerry's QNX will work with Microsoft on software-defined vehicles or SDVs
• QNX's software development platform will be available on Microsoft Azure
• This should shorten the time-to-market, enabling products to be delivered faster

BlackBerry, once synonymous with the business smartphone, is now a major automotive technology player through its QNX platform, and has now revealed a tie-up with Microsoft to revolutionize the development of Software-Defined Vehicles (SDVs).

The company announced its Real-Time Operating System (RTOS) division is collaborating with Microsoft in a bid to streamline the development process for automakers.

The partnership will integrate BlackBerry's software tools into the Microsoft Azure cloud platform, with the QNX Software Development Platform (SDP) 8.0 giving developers an environment for building, testing, and refining automotive software.

QNX Software on Microsoft Azure

Here, the cloud’s scalability should accelerate a typically extended development cycle. Early testing and the virtual validation of software can resolve issues much sooner, improving the time-to-market for software-defined vehicles.

The partnership will later extend to the QNX Hypervisor virtual machine software and QNX Cabin reference architecture, as BlackBerry looks to develop next-generation, hardware agnostic, and cloud-based automotive and Internet of Things (IoT) applications.

“Together with Microsoft, we are looking at innovative ways to bring Microsoft Azure AI in to support cutting-edge technology early in the development cycle," said John Wall, Chief Operating Officer and Head of Product, Engineering and Services at QNX.

“This approach accelerates development timelines and minimizes risks, allowing automakers to leverage the scale and development velocity inherent to the cloud. Together, we're paving the way for the next generation of connected and autonomous vehicles," he added.

Dayan Rodriguez, Corporate Vice President of Manufacturing & Mobility at Microsoft said the collaboration, "bolsters our commitment to empower the automotive industry with cutting-edge technology to accelerate innovation. By combining the strengths of Microsoft Azure and QNX, we’re enabling OEMs to deliver on the promise of the Software-Defined Vehicle.”

You might also like

• These are the best dash cam for every budget
• Here are the best cloud document storage
• The 9 best Android Automotive apps to upgrade your driving experience in 2025

• Unit 42 says phishing campaign targeted automotive, chemical, and industrial compound manufacturing industries
• More than 20,000 victims were successfully targeted
• The campaign has been disrupted, but users should still be on their guard

Hackers of potentially Russian or Ukrainian origin have been targeting UK and EU organizations in the automotive, chemical, and industrial compound manufacturing industries with advanced phishing threats, experts have warned.

A report from Unit 42, Palo Alto Networks’ cybersecurity arm, claims to have observed a campaign that started in June 2024, and was still active as of September. The goal of the campaign was to grab people’s Microsoft Azure cloud accounts, and steal any sensitive information found there.

The crooks would either send a Docusign-enabled PDF file, or an embedded HTML link, which would redirect the victims to a HubSpot Free Form Builder link. That link would usually invite the reader to “View Document on Microsoft Secured Cloud,” where the victims would be asked to provide their Microsoft Azure login credentials.

Bulletproof hosting

The majority of the victims are located in Europe (mostly Germany), and the UK. Roughly 20,000 users were “successfully targeted”, the researchers said, adding that at least in a few cases, the victims provided the attackers with login credentials: "We verified that the phishing campaign did make several attempts to connect to the victims' Microsoft Azure cloud infrastructure," the researchers said in their writeup.

Besides using custom phishing lures, with organization-specific branding and email formats, the crooks also went for targeted redirections using URLs designed to look like the victim organization’s domain. Furthermore, the miscreants used bulletproof VPS hosts, and reused their phishing infrastructure for multiple operations. Most of the phishing pages were hosted on .buzz domains.

At press time, most of the attack infrastructure was pulled offline - Unit 42 said it worked together with HubSpot to address the abuse of the platform, and engaged with compromised organizations to provide recovery resources. Since most phishing servers are now offline, the researchers said the disruption efforts were effective.

Via The Register

You might also like

• Popular astrology app leak exposes data on millions of users — find out if you're affected
• Here's a list of the best antivirus
• These are the best endpoint protection tools right now

Microsoft has released new intelligence claiming Iranian state-sponsored threat actor Peach Sandstorm is using a custom-built backdoor and password spraying attacks for intelligence operations on satellite communications.

The backdoor, named ‘Tickler’ by Microsoft Threat Intelligence, is a specialized multi-stage malware used to compromise target organizations, before moving laterally to gather intelligence using Server Message Block (SMB), remote monitoring and management (RMM) tools, and Active Directory (AD) snapshots.

Tickler has also been used to target oil and gas, and both state and federal level governments in the US and UAE.

Satellite Tickler

Microsoft's Threat Intelligence team says Peach Sandstorm has been observed using password spraying attacks to compromise accounts belonging to target organizations in the education, defense, space, and government sectors.

By compromising accounts in the education sector, Peach Sandstorm would use newly created or existing Azure student subscriptions to host command-and-control (C2) infrastructure. Through this C2 infrastructure, the group would then target organizations within the government, defense and space sectors to gather intelligence on satellite communications equipment.

Two versions of Tickler have been identified by Microsoft. The first was found within a file named ‘Network Security.zip’ alongside a pair of decoy PDF documents. The actual Tickler malware used the same file name as one of the benign PDFs, but was actually an executable with the suffix ‘.pdf.exe’. When launched, the executable file collects network information from the host device by decrypting kernell32.dll, and sends this information to the C2 infrastructure.

The second version functions in exactly the same way as the first, but is also able to download additional malware from the C2 infrastructure to deploy on the host device, allowing for DLL sideloading to establish a backdoor, from which the attackers can run numerous commands to delete files, execute commands, and both download and upload files from the C2 infrastructure.

As an Iranian state-sponsored threat actor, Peach Sandstorm is likely to be operating on the behalf of the Iranian Islamic Revolutionary Guard Corps (IRGC) to further the gathering of intelligence in line with Iranian state interests.

In order to mitigate the exploitation of Azure infrastructure by threat actors using compromised accounts, Microsoft began enforcing multi-factor authentication by default for all Azure administrators from July 2024, before rolling out MFA to all Azure accounts from October 2024.

More from TechRadar Pro

• These are the best authenticator apps
• Google says it has found Iranian hackers hitting top US presidential election targets
• Take a look at the best cloud firewalls

Microsoft has announced all Azure sign-ins will soon have to use multi-factor authentication (MFA) to enhance account security.

The authentication methods offered by the Azure platform include mobile app push notifications, biometrics, one-time passcodes or passkeys using Microsoft Authenticator.

FIDO2 security keys, certificate based authentication, and SMS or voice approval authentication methods are also available.

Azure authentication

The rollout of Azure MFA will occur in two phases. The first phase will require MFA to be used to sign in to the Azure portal, Microsoft Entra admin center, and the Intune admin center, with all Entra global admins receiving a 60-day notice by email and Azure Service Health Notifications. Phase 1 will begin in October 2024.

Phase 2 is expected to start in early 2025, which will require all sign-ins on Azure CLI, Azure Powershell, Azure mobile app, and Infrastructure as Code (IaC) tools. If your organization already has an MFA sign-in policy such as Microsoft defaults or a Conditional Access policy that utilizes MFA, users will not see any change to their log in.

The rollout is part of Microsoft’s Secure Future Initiative (SFI), with part of this program aimed at providing additional phishing resistance through MFA to all user accounts. Last year, China-based state-sponsored hackers managed to hack into a number of email accounts belonging to US government officials.

“Our goal is to deliver a low-friction experience for legitimate customers while ensuring robust security measures are in place. We encourage all customers to begin planning for compliance as soon as possible to avoid any business interruptions,” Microsoft said in a statement.

Via InfosecurityMagazine.

More from TechRadar Pro

• These are the best password managers
• A bank wire transfer scam cost this company millions — here's how you can stay safe
• Take a look at the best antivirus options around today

Despite OpenAI’s recent announcement that it would leave the Chinese market by enforcing its policy to block users in nations outside of officially supported markets, many Chinese customers may still be able to use the GPT-powered AI chatbot.

Although the likes of Alibaba, Baidu and Tencent have been rushing to fill the void left by California-based OpenAI, fans of ChatGPT can still use a very similar service through a Microsoft-enabled loophole.

Those with access to Azure cloud services will still be able to use GPT-powered generative AI functionalities in China, even though the startup responsible for GPT, OpenAI, has formally left the country.

Chinese customers can still use Microsoft’s ChatGPT

Although ChatGPT was blocked in China, developers valued its productivity benefits and circumvented restrictions by using VPNs. However, OpenAI’s abrupt decision to enforce the restriction, effective July 9, left many developers and companies relying on the technology scrambling to find a suitable alternative. Subsequently, Chinese LLM developers launched their own aggressive campaigns to attract ex-ChatGPT users.

However, those already familiar with the workings of ChatGPT will still be able to use it, just through a Microsoft-enabled loophole. Operated as a joint venture with domestic tech firm 21Vianet, Microsoft Azure China continues to provide OpenAI models in the country.

Microsoft China even encouraged developers to transition to Azure OpenAI via its official WeChat account (via Tom’s Hardware).

Three Azure China customers also confirmed to The Information that they still have uninterrupted access to OpenAI’s models through Microsoft’s platform.

While access currently remains available to Chinese customers, a bill aimed at simplifying the process of blocking US companies from selling AI technology to China is currently under consideration. If enacted, it could prevent Microsoft from being able to distribute OpenAI models to customers in China altogether, handing back its portion of the market to Chinese-owned rival companies.

More from TechRadar Pro

• Check out the best AI tools and best AI writers
• China is planning a huge national compute boost by 2025
• We’ve rounded up a list of all the very best cloud hosting providers

Microsoft has confirmed it will start enforcing multi-factor authentication (MFA) for all Azure administrators within a matter of weeks.

In a blog post, Principal Product Manager for Microsoft Azure, Naj Shahid, outlined how the move would help address a number of customer concerns.

The rollout will start in July 2024, and will first be available to Azure admins. After that, similar rollouts will happen for CLI, PowerShell, and Terraform. Users will be notified beforehand via email.

Defending the premises

"Service principals, managed identities, workload identities, and similar token-based accounts used for automation are excluded," according to Shahid. "Microsoft is still gathering customer input for certain scenarios such as break-glass accounts and other special recovery processes." 

"Students, guest users and other end-users will only be affected if they are signing into Azure portal, CLI, PowerShell or Terraform to administer Azure resources. This enforcement policy does not extend to apps, websites or services hosted on Azure. The authentication policy for those will still be controlled by the app, website or service owners."

MFA adds a second authentication layer, besides the password, to high-value accounts. It generally comes in the form of a time-based code that is generated by a MFA tool, such as an authenticator app, or a physical token. 

These days, MFA is considered an industry standard in terms of cybersecurity and comes highly recommended for customers, as it successfully repels a vast majority of cyberattacks, and makes phishing for passwords extremely difficult.

Some phishing kits allow threat actors to steal multi-factor authentication codes, too, but the process is a lot more cumbersome, and hackers are a lot easier to oust, compared to accounts without MFA.

More from TechRadar Pro

• Microsoft wants to take any MFA and 2FA worries out of your hands
• Here's a list of the best firewalls today
• These are the best endpoint protection tools right now

Microsoft has fixed a flaw that leaked company files and employee data online.

Researchers from cybersecurity firm SOCRadar discovered a server on Microsoft Azure, the company's cloud computing platform, that was open to the public and contained information pertaining to its Bing search engine.

Passwords and other credentials belonging to employees were also exposed, which could have given hackers a golden opportunity to breach other databases belonging to the tech giant.

Same mistakes

The database the researchers found was not protected with a password, allowing for anyone to access it. 

One of the researchers told TechCrunch that the data within could have allowed bad actors to locate other areas where Microsoft holds valuable data, which could have led to compromises of its services.

Microsoft was notified by SOCRadar of the issue on February 6 and Microsoft applied a fix on March 5. TechCrunch reports that it is not known how long the server was exposed for, or it anyone else had discovered the leak.

Leaking databases are a persistence problem, with many organizations failing to secure them properly. For example, in February, Zenlayer, a network services provider, failed to password-protect a database containing sensitive customer and company information.

Microsoft itself has also been guilty of the similar lapses in the past. In July last year, Chinese hackers stole a signing key to target email accounts belonging to government and business organizations, with the company conceding that it did not know how they had managed to acquire them.

MORE FROM TECHRADAR PRO

• This is the best endpoint protection for business and organizations
• One of the biggest data leaks ever has just been revealed - here's what to do if you've been hit
• A new Microsoft Azure hacking campaign is targeting high-end executives

In the wake of similar moves by rival companies Google Cloud and Amazon Web Services (AWS), Microsoft has announced that it will now be scrapping certain egress fees for its Azure cloud computing service.

The news was confirmed in a company blog post, which states: “We support customer choice, including the choice to migrate your data away from Azure.”

The company emphasized that Azure now offers free egress for customers transitioning their data out of the platform, be it to another rival cloud provider or an on-premises data center.

Azure removes cloud egress fees

As with its rival companies, the decision isn’t one of goodwill. Rather, it aligns with the impending enforcement of provisions within the European Data Act, aimed at dismantling contractual terms that prevent and hinder competition through vendor lock-in.

Microsoft is already under a lot of pressure across Europe, with the UK’s CMA investigating its egress fees and other antitrust cases continue to develop, including unfair licensing that has led Microsoft to have dominance over the market.

However, although the impending regulations are set to affect business across Europe, like Google and Amazon, Microsoft has confirmed that it will remove egress fees for “all Azure customers globally and from any Azure region.”

Up until now, customers have been limited to the company’s 100GB-per-month free limit, and in a way, that will still apply. Rather than totally scrapping the fees, customers seeking to migrate must obtain credit to cover the cost of exiting. Microsoft’s guidance points to this support page for further guidance.

The announcement reads: “Azure Support will apply the credit when the data transfer process is complete and all Azure subscriptions associated to the account have been canceled.”

Microsoft declined to add any further comment.

More from TechRadar Pro

• These are the best cloud hosting providers
• Microsoft issued cloud pricing ultimatum by EU states
• Store your data in the best cloud storage or the best cloud backup tools

Google has announced the launch of Security Command Center Enterprise (SCC Enterprise), a new cloud risk management solution offering what it says is an industry-first combination of cloud security and enterprise security operations.

And to wrap it all up, this security solution is backed up with Mandiant threat intelligence and generative AI to provide clarity and insight throughout the entire cloud security lifecycle.

In a company blog post, Google Cloud says cloud-native application protection platforms (CNAPPs) don’t have all of the capabilities necessary to provide the best protection against the latest threats, which is why it is looking to introduce such a security command center for enterprise.

SCC Enterprise supercharged by Mandiant Hunt

By combining Mandiant Threat intelligence with modern SecOps capabilities, SCC Enterprise provides a rapid response solution to cloud security issues by utilizing “SIEM-powered visibility and SOAR-driven accountability,” Suni Potti, VP/GM, Google Cloud Security said on the blog.

“Security teams can get a single view of their posture controls, active threats, cloud identities, data, and more, while integrating remediation and issue accountability into the end-to-end workflows of a converged cloud risk management platform.”

Google’s security fabric is the powerhouse behind SCC Enterprise, giving insight into the data visualizations provided by the “planet-scale data lake that ingests and analyzes the volume of cloud data needed to build dynamic security graphs and to understand the complex relationships in multicloud environments.”

For those with less-experience in managing a multi-cloud security environment, Google has included generative AI to simplify the complex security issues and terminology that are encountered when navigating SCC Enterprise.

The Mandiant Hunt threat intelligence expertise can be accessed on demand to enhance the skills and tools available to security teams, giving your team's access to the latest threats to cloud environments and the best techniques for prevention and remediation.

The Enterprise level solution builds on Google’s pre existing Command Center with additional capabilities for securing Google Cloud, AWS, and Microsoft Azure environments, including vulnerability and security posture management, cloud infrastructure specific threat detection technology, integrated workflows and much more.

More from TechRadar Pro

• Google's presentation software has a new Gemini AI powered feature that is basically magic
• These are the best cloud storage and the best cloud computing services
• Thousands of Fortinet devices could face attack following security issue
  

We list the best IaaS providers, to make it simple and easy to provide IT infrastructure services based in the cloud.

In the internet’s early days, it was much more difficult to build your own web service. You had to buy and maintain your own servers or enter binding long-term contracts with data center operators. This arrangement was inconvenient for most businesses, and the costs were high, meaning people weren’t too willing to take risks. However, cloud computing providers changed the game.

Cloud computing platforms operate servers and rent them to customers on demand. There’s no need to sign any long-term contract, as you only pay for what you use. Anyone can now host their cloud-based app conveniently and affordably. This arrangement is known as Infrastructure as a Service (IaaS).

You can rent a server from an IaaS provider and run any operating system or application you want on it. You can decide which locations to rent from to ensure proximity to your users, which, in turn, delivers faster speed.

We tested different IaaS providers to identify the best ones. We narrowed our list based on important factors including pricing, features, performance, ease of use, customer support, etc. The platforms we selected have their pros and cons, but are, overall, the best ones you can choose to host your websites and apps.

We've also listed the best PaaS provider.

The best IaaS provider of 2025 in full:

Best IaaS provider for reliability

Amazon Web Services (AWS) is the world’s biggest cloud computing provider. The company operates over 125 data centers across the globe, delivering fast performance and reliability with this massive server network.

Amazon EC2 allows you to rent virtual computers from AWS to run your applications. You can choose your operating system, storage space, processor, and many other variables. AWS boasts of being the only cloud provider with 400 Gbps Ethernet networking, meaning unprecedented speeds. It's also the first major cloud provider supporting AMD, Intel, and Arm processors, so you can choose the one that best fits

your app.

EC2 is pay-as-you-go, so you only pay for the resources you use. However, the pricing structure isn’t straightforward, which makes it difficult to calculate what your final costs will be. AWS is known for being affordable at low usage but expensive if you use significant computing resources.

EC2 has advanced security features to keep your sensitive data safe. Overall, it offers one of the broadest feature sets you can find in an IaaS provider; this is an advantage, but means that you may face difficulty navigating through the numerous features. EC2 is infamous for having a complex interface, and you may need to hire dedicated AWS experts that don’t come cheap.

• ^ Back to the top

Best IaaS provider for security

Azure is a cloud computing service offered by Microsoft, the tech giant best known for its Windows operating system. Azure is the second-biggest cloud provider after AWS, with a network of over 100 data centers globally.

You can rent virtual machines on Azure running either the Windows or Linux operating system. You can deploy different types of software on these servers, e.g., SQL server, and SAP ERP.

Azure has very advanced security to protect you from malicious threats and guarantee secure network traffic. It has thousands of cybersecurity experts constantly identifying and thwarting potential threats.

This platform is very scalable. You can start from one virtual machine and go as high as you want as long as you pay. Azure virtual machines are known to be affordable, especially at a high scale. It’s pay-as-you-go, so you only pay for the resources you consume, just like with other cloud providers.

A major complaint about Azure is that it has a pretty complex interface. There’s a steep learning curve, but you can get used to it with time.

• ^ Back to the top

Best IaaS provider for performance

Google is the tech giant best known for its eponymous search engine. However, it has many other products, including the Google Cloud system. Compute Engine is an IaaS service offered as part of Google Cloud. It lets you deploy and run virtual machines on Google's powerful data center infrastructure.

You can easily create and manage your fleet of virtual machines via Google Compute Engine; the machines ran either run the Windows or Linux operating system. You can live-migrate virtual machines between different hosts without rebooting, meaning your applications will continue running even while the system needs maintenance.

You can create Confidential virtual machines that encrypt data while in use. Google says the Confidential VMs don’t compromise on performance despite the additional encryption. You can spin up VMs with up to 11.5 TB of RAM and 60 virtual centralized processing units (vCPUs), signaling Google Cloud’s high performance.

Google Compute Engine is known for its very high performance and cost-effectiveness. However, this platform has a steep learning curve, so you might it difficult to use at first.

• ^ Back to the top

Best IaaS provider for affordability

DigitalOcean is a low-cost cloud computing provider that’s ideal for small businesses and individual use. You can rent virtual machines called Droplets and add more when required. The platform promises 99.99% uptime and advanced security features to keep your data safe.

DigitalOcean has very straightforward pricing. Droplets start from $4 per month, which is affordable, and you can deploy them in seconds. You can run any type of app you want with the Droplets.

DigitalOcean also offers Kubernetes clusters starting from $12 per month. You can deploy these clusters without much complexity.

The best thing about DigitalOcean is that it’s much more affordable than the competition. It’s also very easy to use, with an intuitive interface that you’ll likely find easy to navigate. Customer support is great, as often highlighted in user reviews.

However, DigitalOcean doesn’t offer as many complex features as you’ll find on rivals like Compute Engine and Amazon EC2.

• ^ Back to the top

Best IaaS provider for Linux servers

Linode is a cloud hosting provider focused on Linux-powered virtual machines. You can rent powerful CPUs and servers running the Linux operating system, which is preferable for some people.

This platform is self-serve, so you can easily sign up and spin up the servers you need. You can rent CPUs with different levels of storage, RAM, and capacity. For instance, you can rent CPUs that need to be up throughout the day.

The virtual machines you rent can be dedicated or shared. The former is more expensive but gives better performance and speed. The latter is slower but still reliable to run your applications.

Linode is very secure, with bundled distributed denial of service (DDOS) protection and cloud firewalls. You can access many other extra features such as the Domain Name System (DNS) manager and a cloud management app.

Linode offers great value for money and has excellent customer support, as user reviews often point out. The drawback is that it’s limited to Linux and doesn’t offer as many features as you’ll find on rival IaaS providers.

• ^ Back to the top

Best IaaS provider for enterprise

Founded 1996 by a group of software developers, Kamatera provides enterprise-grade cloud services and is used by everyone from international enterprises to SMBs.

Products offered include cloud servers, cloud desktops, virtual private clouds, cloud firewalls, load balancers, block storage, and reseller hosting. For each of the products you can choose to configure your server to meet the demands of your enterprise across a wide range of individually customizable options, offering up to 32 vCPUs, 128GB RAM, 4TB of high speed SSD storage, and a range of hosting locations across Europe, North America, Asia, and the Middle East.

You can choose from a selection of different pre-installed Linux distributions which includes Ubuntu, centOS, Debian and FreeBSD and are included in the standard pricing. Desktop and Server versions of Windows are also available, but they do come at an additional cost that varies depending on the version, however you have the option to manually install any OS of your choice and use any licences you currently have.

In addition, there are 4 types of server available to tailor the performance of your server to your needs that range from highly available, but cheaper, non dedicated resources, to fully dedicated resources at a greater cost. Enterprises can also configure their servers with extended daily backups to an external storage array, and a fully managed service to help with OS and system configuration. Alongside the ability to fully customizable a server, Kamatera also provides 3 standard configurations where all you need to select is the location to deploy the server, and the OS you’d like to use.

The cheapest option offers 1 vCPU, 1GB RAM and 20GB SSD storage, while the more expensive package offers 2 vCPUs, 2GB RAM and 30GB SSD storage. All packages, whether customized or standard configurations come with 5TB of internet traffic per month, with additional charges for any bandwidth over the limit.

Kamatera facilities have 24/7 surveillance and top level security to ensure that your data is safe and secure. The power infrastructure is designed to be failure-proof and all servers are connected to conditioned UPSs, N + 1 redundant instantaneous failover, and independent generators on site to power the system in the event disaster strikes to prevent any interruption to your services to maintain their promised uptime SLA of 99.95%.

All servers are built using the latest Gold and Ice Lake Intel Xeon processors offering speeds of 2.7GHz+, storage arrays comprise of SSDs or use SSD accelerators to provide lightning fast storage speeds, and each server is connected to their network and router backbones to offer 10Gbit/s speeds per virtual server with network infrastructure backed by tier-1 providers worldwide.

Kamatera offers all enterprise grade customer support that operates 24/7, 365 days per year, across all time zones. Customers can contact the support team using their preferred choice of email, live chat or phone. For large enterprises, Kamatera can also offer a personalized service that includes a dedicated cloud administrator to provide continuous monitoring and management of your servers.

If you’d like to try before you buy, Kamatera offers a 30 day free trial that allows you to create and deploy a server configuration with a value up to $100 and includes 5TB of traffic (reduced to 1TB if deployed in Hong Kong)

• ^ Back to the top

We've also highlighted the best business intelligence tools.

Best IaaS provider FAQs

How to choose an Iaas provider 

Cost

Cost is the main consideration when choosing any service. You should choose an IaaS provider that you can afford in the long term to avoid troubles. Different providers have different pricing structures, so study them to identify any potential problems. For instance, some hosts are affordable in the beginning but become costlier as you scale. Always read the fine print to have the full picture of how the pricing works.

Performance and Reliability

Your IaaS provider must be a reliable one. You don’t want something that goes down frequently and causes you to lose money. The servers you rent should also deliver solid performance and make your website or online app load as quickly as possible.

The larger the data center network of your cloud provider, the better their reliability and redundancy. Performance depends on what type of servers they use and the proximity to your users.

Customer Support

Nothing is perfect including IaaS providers. Things can go wrong at any time, and you’ll likely need customer support. You should be able to contact a human support representative from your IaaS host to help you solve your issues. The host should also provide complementary support resources such as FAQs, user guides, video tutorials, and extensive documentation.

What are the essential features of an IaaS provider? 

1. Pay-As-You-Go

The main benefit of IaaS providers is that you don’t have to sign any long-term contracts to rent computing resources. You can just select the ones you want and pay as you go. You only pay for what you use, which makes it convenient and more affordable.

2. Scalability

You should be able to scale your computing needs with ease. Your IaaS provider should handle your increasing data storage and computing needs without hassle. You just need to select the additional resources you need and deploy them.

3. Security

You’ll be uploading sensitive data to your IaaS provider, so it should have sophisticated security features to prevent data leaks and theft. Losing customer data could lead to costly fines and lawsuits and losing sensitive business data could make you lose your competitive advantage.

4. Virtualization

Virtualization means creating virtual versions of computers or servers. One server can host several virtual machines, with each machine acting like its own separate server with distinct features. Your IaaS provider should let you deploy virtualized computing instances to support your app.

More on IaaS

Investing in your own IT infrastructure is a significant cost, regardless as to whether you're a startup or an expanding business.

It's not just the servers, but all aspects of your network that you will need to buy, and for some companies this can be prohibitive to growth. Even worse is that after a few years this hardware will need replacing as an additional cost.

However, advances in cloud computing mean that an alternative that is now available: Platform as a Service (PaaS), in which a supplier already has their own servers and networks, and you simply rent the resources you need.

The big advantage is that pricing is much less than if you bought your own hardware, making it much easier for small companies to set up an extensive and professional IT infrastructure at the fraction of the cost.

Additionally, because you are only paying for what you need, there's no over-spend required for redundant systems, as IaaS is fully scaleable for as and when you need it.

How to choose an Iaas provider 

Cost

Cost is the main consideration when choosing any service. You should choose an IaaS provider that you can afford in the long term to avoid troubles. Different providers have different pricing structures, so study them to identify any potential problems. For instance, some hosts are affordable in the beginning but become costlier as you scale. Always read the fine print to have the full picture of how the pricing works.

Performance and Reliability

Your IaaS provider must be a reliable one. You don’t want something that goes down frequently and causes you to lose money. The servers you rent should also deliver solid performance and make your website or online app load as quickly as possible.

The larger the data center network of your cloud provider, the better their reliability and redundancy. Performance depends on what type of servers they use and the proximity to your users.

Customer Support

Nothing is perfect including IaaS providers. Things can go wrong at any time, and you’ll likely need customer support. You should be able to contact a human support representative from your IaaS host to help you solve your issues. The host should also provide complementary support resources such as FAQs, user guides, video tutorials, and extensive documentation.

How we tested the best IaaS provider

We test by evaluating numerous factors. To start with, we look at the feature set, the range of tools available, and what size of businesses this would be ideal for. We consider how easy the setup is, the simplicity of the interface, and whether there's sufficient documentation and tutorials for users to utilize necessary options optimally.

We assess how well the service integrates with other relevant apps, and check the overall scalability of the service. We also analyze whether there are collaboration features for multiple users, and lastly, we judge the quality of the customer service and the different pricing plans available.

Read more on how we test, rate, and review products on TechRadar.

Get in touch

• Want to find out about commercial or marketing opportunities? Click here
• Out of date info, errors, complaints or broken links? Give us a nudge
• Got a suggestion for a product or service provider? Message us directly

• You've reached the end of the page. Jump back up to the top ^

The best Platform as a Service (PaaS) providers make it simple and easy to run your apps using Platform as a Service.

It has become easier than ever to build and develop and host applications, thanks to platform as a service (PaaS) providers. These providers offer the tools that you need to develop, run, and maintain apps on the cloud.

It’s essential not to confuse Infrastructure as a Service (IaaS) for PaaS. They are related but not the same thing. IaaS deals with providing the hardware infrastructure needed to host online services, while PaaS deals with providing the development tools needed to build these services.

You can access PaaS in three major ways. The first and most popular is a dedicated service from a cloud computing provider. You control the software aspects, including the operating system and programming languages, while the cloud provider provides the servers and networking equipment to hold the online services.

The second way to access PaaS is as a private service behind a firewall and the third is as software deployed on public infrastructure. The primary aim of a PaaS provider is to simplify the application development process by taking care of the complex aspects while you handle the easier aspects.

We tested different PaaS providers to identify the best ones. We selected them based on important factors including pricing, performance, features, customer service, etc.

We've also highlighted the best IaaS providers.

The best PaaS providers of 2026 in full:

Best PaaS provider for reliability

Amazon Web Services (AWS) is the world’s biggest cloud computing provider, with a network of over 200 data centers across the globe. Elastic Beanstalk is a PaaS service built atop AWS.

Elastic Beanstalk provides a platform to let you build and deploy web-based applications easily. The AWS management console is famous for being easy to use and navigate, so you shouldn't face difficulty trying to work your way around this platform. Deploying your apps is easy and as your app becomes more complex, you can upgrade the infrastructure.

Elastic Beanstalk operates on a pay-as-you-go model. The platform itself doesn't incur costs. Instead, you only pay for the AWS infrastructure (e.g., S3 buckets and EC2 cloud instances) that your app consumes. There are no minimum fees or upfront costs needed. AWS has a reputation for being economical on the low end and expensive if your app has high traffic.

Elastic Beanstalk relies on AWS infrastructure, which is very reliable and redundant. AWS guarantees a 99.99% uptime. The main drawback we observed with this platform is that its app debugging and monitoring tools were difficult to use.

• ^ Back to the top

Best PaaS provider for security

Azure is a cloud computing service offered by Microsoft, the technology giant best known for its Windows operating system. Microsoft bundles its PaaS and IaaS services together under Azure.

You can use the Azure App Service to create and deploy your apps easily. This platform lets you build web-based apps for personal computers and mobile phones. It supports various programming languages including Python, Java, Node.js, Core, and Microsoft's own .NET. 

Azure App Service offers a fully managed platforms that takes care of all infrastructural needs while you focus on programming your app. Azure is famous for its sophisticated security features. It meets rigorous compliance standards, such as SOC and PCI. Microsoft employs thousands of cybersecurity experts dedicated to keeping everything about Azure secure.

Azure has a robust physical infrastructure, with over 200 data centers across the globe, so expect near-perfect uptime for your applications. The main complaints we can point out about Azure is that the interface is difficult to understand, especially for a first-time user. It also has a complicated pricing structure that makes it difficult to estimate your final app hosting costs. 

• ^ Back to the top

Best PaaS provider for performance

Google is best known as a search engine but the company offers many other services including the Google Cloud computing service. Google App Engine is a PaaS built atop Google Cloud. It provides tools for developers to build and host web applications on Google’s powerful data center infrastructure.

Google App Engine supports many popular programming languages including Java, C#, Python, PHP, Go, and Node.js. It offers a fully managed environment, managing all infrastructural needs while you focus on writing code for your app.

App Engine makes it easy to monitor your app's performance; you can use the Cloud Monitoring and Cloud Logging tools to monitor the performance and the Cloud Debugger to identify bugs and fix them quickly. You can easily host different versions of your apps on App Engine and create different development, testing, and production environments for them from a single account.

You can safeguard your app using a firewall and setting custom access rules. Google’s physical server infrastructure is very powerful, so you shouldn’t have problems with scaling your app. However, a major drawback of using App Engine is that it’s more expensive than rival tools.

• ^ Back to the top

Best PaaS provider for ease of use

Heroku is a simple platform as a service (PaaS) provider. It makes building and deploying web applications very easy. The platform handles the hard part of maintaining the infrastructure, while you handle the easier parts of programming your app.

Heroku supports various programming languages, including Ruby, Java, PHP, Python, Scala, Clojure, and Go. You can also use several unofficially supported languages via third-party Buildpacks.

The best thing about Heroku is its simplicity. It’s one of the easiest platforms for any developer to understand. For instance, you can create a new server in about 10 seconds by using a simple code on the command line. Check Heroku reviews, and you’ll often see developer experience as the main highlight.

Heroku is relatively affordable if your app doesn’t have much traffic. However, it gets expensive as the traffic grows. Heroku is best used for small projects, as it can be challenging work with if your app gets high traffic.

• ^ Back to the top

Best PaaS provider for affordability

DigitalOcean is a niche cloud computing provider. Its primary appeal is being a low-cost alternative to the more popular platforms such as AWS, Azure, and Google Cloud.

App Platform is DigitalOcean’s PaaS solution. It offers a fully managed solution to build and deploy your apps. DigitalOcean has an intuitive interface that you'll likely enjoy using.

App Platform secures your apps automatically. It automatically renews your SSL certificates and protects them from distributed denial of service (DDOS) attacks, so you don't have to worry about that.

App Platform supports many programming languages including Python, Node.js, PHP, Ruby, and Go. If your app was built in an unspported language, you can simply provide a Container Image, and DigitalOcean can still run it.

App Platform has a free plan, which lets you build static sites. To build apps, you’ll need a premium plan which starts at just $5 per month. You can choose to use either a shared or dedicated CPU for your app, but the latter costs more.

• ^ Back to the top

We've also listed the best cloud orchestration software.

Best PaaS provider FAQs

How to choose a PaaS provider 

1. Programming languages

What programming languages do your PaaS provider support? The provider should be "polyglot", meaning it supports multiple languages and lets you choose the one you're familiar with, such as Java, Javascript, C++, Perl, PHB, Ruby, etc.

2. Supported frameworks

A software framework refers to the structure you use to build an online service, e.g., Ruby on Rails for Ruby, Django for Python, and Laravel for PHP. Your PaaS provider should support frameworks that you're familiar with to make it easy for you to develop apps. 

3. Scalability

Your PaaS platform should allow you to scale your app with little hassles. As you add users and your data needs grow, it should be able to keep up without reducing the quality of service.

4. Run-time environment

Run-time environment (RTE) refers to the operating system that allows your application to run. Common examples include Apache Tomcat, Jetty, and IIS. Your PaaS provider should support many run-time environments and give you a broad choice for app development.

5. Cost

Cost is a primary consideration when choosing any platform. You should choose a PaaS provider that you can easily afford in the long term. Check the pricing pages and read the fine print to see if there are any hidden costs.

Features of a PaaS provider 

1. Web-based development environment

Your PaaS platform should offer a browser-based development environment where you can write code and deploy it in real time. You should be able to see the effects of your code immediately after deploying it. 

2. Monitoring and analytics

Your PaaS should allow you to analyze your app’s usage. You should get a breakdown of what resources your app is consuming and how much each one contributes to your final bill. This will make it easy to optimize your app development for lower costs.

3. Database

It should provide database software to store data for your applications. It should also allow you to configure and optimize your database for effective results. 

4. Middleware

Middleware refers to software that lies between an operating system and the apps running on it. It includes web servers, content management systems, data integration tools, application programming interfaces (APIs), etc.

Your PaaS platform should provide middleware services to make it easy to develop your apps. Otherwise, your developers will have to spend more time and effort creating these services by themselves.

How we tested the best PaaS provider

We test by evaluating numerous factors. To start with, we look at the feature set, the range of tools available, and what size of businesses this would be ideal for. We consider how easy the setup is, the simplicity of the interface, and whether there's sufficient documentation and tutorials for users to utilize necessary options optimally. 

We assess how well the service integrates with other relevant apps, and check the overall scalability of the service. We also analyze whether there are collaboration features for multiple users, and lastly, we judge the quality of the customer service and the different pricing plans available.

Read more on how we test, rate, and review products on TechRadar.

Get in touch

• Want to find out about commercial or marketing opportunities? Click here
• Out of date info, errors, complaints or broken links? Give us a nudge
• Got a suggestion for a product or service provider? Message us directly

• You've reached the end of the page. Jump back up to the top ^

Microsoft is introducing AI to the telecommunications industry to help protect people from spam and scams.

Announced at the Mobile World Congress in Barcelona this week, Azure Operator Call Protection is now available on public preview.

The AI tool requires no download or installation, and provides endpoint protection directly through the network.

AI call monitoring

Azure Operator Protection uses real-time analysis of phone calls to determine if the content is suspicious, and then provides an alert to the call handler if they are at risk of being scammed. In the UK, BT Group is trialing the product to help protect its customers, particularly those who are more vulnerable to scams.

Scam calls cost businesses and individuals billions of dollars per year, with the FTC estimating that the total cost of such calls in the US amounted to $850 million in 2023 alone.

Spam and scam calls are expected to become more prevalent in the run up to a number of important elections around the world, with a recent example featuring a spoof of President Biden, telling New Hampshire constituents not to vote in a state election.

A number of top tech and social media companies have recently signed a voluntary agreement to identify, label, and combat deepfakes that could be used for political interference or to mislead voters.

More from TechRadar Pro

• Most data breaches on enterprise attack the supply chain
• Github Copilot Enterprise AI coding assistant for businesses hits general availability, but it'll cost you
• Google ramps up its Microsoft criticism — a cloud war could be on the horizon
  

Google Cloud has taken a swing at Microsoft, claiming its rival company is taking steps to secure a monopoly on emerging technologies such as generative AI.

Microsoft controls a significant portion of the cloud computing market, and has recently begun a collaboration with OpenAI to enhance their development of all things AI.

This has resulted in increasing scrutiny from regulatory bodies in the US, UK and the European Union.

Microsoft is “creating a whole walled garden”

Google Cloud President Amit Zavery recently expressed his concern about Microsoft’s grip on the market, saying, “We worry about Microsoft wanting to flex their decade-long practices where they had a lot of monopoly on the on-premise software before and now they are trying to push that into cloud now.

“So they are creating this whole walled garden, which is completely controlled and owned by Microsoft, and customers who want to do any of this stuff, you have to go to Microsoft only.”

Zavery continued by claiming that Microsoft’s Azure cloud platform forced customers to exclusively use Microsoft products, and that regulators should step in before the cloud giant causes “long-term problems” in other industries.

The Google Cloud President accused Microsoft of flexing its market control by selectively choosing vendors in a manner that can either set them up for success or fall through, allowing Microsoft to choose its competition.

Microsoft’s President Brad Smith took a swing back at Google during the opening day of the Mobile World Congress in Barcelona, saying, “Today, only one company is vertically integrated in a manner that includes every AI layer from chips to a thriving mobile app store.”

Another Microsoft spokesperson also addressed the criticism stating, “We have listened to and work constructively and directly with independent cloud providers to change our licensing terms, addressing their concerns and providing more opportunity for them. Worldwide, more than 100 cloud providers have already taken advantage of these changes.”

Via Reuters

More from TechRadar Pro

• Microsoft's UFS 4.0 storage runs AI faster and optimizes performance during intensive smartphone use
• These are the best cloud storage solutions around
• VMware launches new VeloCloud SASE to help tie together all your edge infrastructure

Microsoft's plans to transition LinkedIn to Azure have been put on hold, according to individuals familiar with the matter (via CNBC).

The project, codenamed ‘Blueshift’, had been announced in 2019, three years after Microsoft acquired the social media and networking platform. Until now, LinkedIn has been using its own data centers, and with the project now seemingly cancelled, it looks like that will continue to be the case.

Some Microsoft workers report having learned about the project’s pause last year, but a new document seen by CNBC details a mutual agreement between Microsoft and LinkedIn to hold off on the migration.

LinkedIn won’t use Azure

The precise reasons behind both companies’ decisions are unclear. LinkedIn CTO Raghu Hiremagalur said in a memo to employees last year: "With the incredible demand Azure is seeing and the growth of our platform, we’ve decided to pause our planned migration of LinkedIn to allocate resources to external Azure customers."

Microsoft Cloud revenue, which includes Azure and other businesses, was up 22.1% in the company’s 2023 fiscal year, which ended June 30. During that period, interest in Microsoft-backed OpenAI generative AI products boomed, in turn knocking up revenue. LinkedIn also saw a healthy growth of 9.6%.

Redmond has also been in a lot of trouble all over the world from various antitrust regulators. Key to many of these cases have been its unfair dominance and anticompetitive business practices, often relating to cloud services. However, Microsoft continues to hold second place, trailing far behind AWS.

A LinkedIn spokesperson commented on the matter: “We are using both Azure to complement our infrastructure needs and further investing in our data centers.”

One of the people familiar with the matter added that LinkedIn is in the process of building another data center to handle its computing needs, suggesting that the platform has no immediate intention of seeing through its Project Blueshift.

More from TechRadar Pro

• Check out the best cloud hosting providers
• LinkedIn is making its platform more accessible with some useful Microsoft tools
• Want to keep data off-prem? Here are the best cloud storage and best cloud backup tools

Microsoft has released a paper for the widely-anticipated glass-based storage technology it's backing to replace the conventional technology that's fitted into the best hard drives and best SSDs out there today.

The 16-page academic paper, presented at the 29th ACM Symposium on Operating Systems Principles, outlines the principles behind the company's plans to build a longlasting and highly efficient storage systems.

Made from quartz glass, the storage units will be primed for use in the cloud – which means Azure customers will be the first to benefit, and likely the only ones to benefit so long as the technology is embryonic in nature.

Project Silica has been years in the making – with Microsoft teaing a prototype as far back as 2019. It's since expanded on its work ahead of designing a system that works in a remarkably similar fashion to the ceramics-based storage that Cerabyte is building.

How does glass-based storage work? 

"This paper presents Silica: the first cloud storage system for archival data underpinned by quartz glass, an extremely resilient media that allows data to be left in situ indefinitely," the authors wrote. 

"The hardware and software of Silica have been co-designed and co-optimized from the media up to the service level with sustainability as a primary objective."

Data is written in a square glass platter with ultrafast femtosecond lasers through voxels. These are permanent modifications to the physical structure of the glass, and allow for multiple bits of data to be written in layers across the surface of the glass. These layers are then stacked vertically in their hundreds.

To read data, they employ polarization microscopy technology to image the platter, while the read drive scans sectors in a Z-pattern. The images are then sent to be processed and decoded, which leans on machine learning model to convert analog signals to digital data.

The medium is suitable for a host of sensitive industries including finance, scientific research and healthcare, due to the secure nature of archival glass storage – meaning organizations in these sectors may be able to withstand ransomware attacks targeting data being stored in the cloud. 

Microsoft is currently exploring how best to configure the physical library in which the glass is stored, in light of Azure cloud storage usage patterns.  

More from TechRadar Pro

• These are the best hard drives out there
• Microsoft is edging closer to making DNA storage a reality
• Video of ceramic storage system prototype surfaces online — 10,000TB cartridges bombarded with laser rays could become mainstream by 2030, making slow hard drives and tapes obsolete

Microsoft has introduced a new hardware upgrade option called Azure Boost designed to improve the performance of virtual machines on its cloud computing platform.

In an explainer, the company outlined how Azure Boost looks to “[offload] server virtualization processes traditionally performed by the hypervisor and host OS onto purpose-built software and hardware.”

Microsoft says that customers can expect a range of benefits including improved networking, revised storage, performance, and security.

Azure Boost – the future of cloud?

Microsoft said that Azure Boost-compatible VM hosts contain the new Microsoft Azure Network Adapter (MANA), which means up to 200 Gbps bandwidth.

Offloading storage operations also sees improvements, with up to 17.3 GBps and 3.8 million IOPS for local storage and up to 12.5 GBps throughput and 650 K IOPS for remote storage.

The cloud provider has confirmed that 17 instance types are already supported, and future Azure virtual machines will also be compatible with Azure Boost.

The explainer confirms that Azure Boost applies to both Linux VMs and Windows VMs, and already, some versions of Ubuntu (20.04 and 22.04 LTS) and Windows Server (2016, 2019, and 2022), among others, have support for the MANA driver (via The Register).

AWS has a similar setup, which it calls Nitro, which underpins the EC2 instances. Like Azure Boost, it promises to offload functions to dedicated hardware and software, in turn improving performance and optimizing cost efficiency.

Whether you’re an AWS customer or an Azure customer preparing to take advantage of Boost, the performance enhancements could allow you to use fewer resources, which is better for both the environment and your business’s bank account.

More from TechRadar Pro

• We’ve rounded up the best cloud hosting
• Microsoft unveils bigger and more powerful Azure VMs
• Check out the best cloud storage and best cloud backup tools

Microsoft has revealed it is launching its own-brand silicon hardware in order to help further the development and use of artificial intelligence in businesses everywhere.

Announced at the company's Microsoft Ignite 2023 event, the new Azure Maia and Azure Cobalt chips put the tech giant fully into the AI hardware arms race.

Microsoft says its new hardware has been built, "with a holistic view of hardware and software systems to optimize performance and price" and form the "last puzzle piece" in its aim to deliver infrastructure systems featuring unparalled optmization for a customer's specific needs.

Microsoft Azure silicon

Specific details on the new hardware were somewhat slim, but AI tools and tasks are unsurprisingly the target for the new releases, with the new Microsoft Azure Maia AI accelerator set to power some of the company's largest internal AI workloads running on Azure, including OpenAI models, Bing, GitHub Copilot and ChatGPT, hopefully producing huge gains in performance and efficiency

The hardware has been built with the help of OpenAI, whose CEO Sam Altman noted that the two companies had collaborated to co-design Azure’s AI infrastructure at every layer for its models and unprecedented training needs.

“We were excited when Microsoft first shared their designs for the Maia chip, and we’ve worked together to refine and test it with our models. Azure’s end-to-end AI architecture, now optimized down to the silicon with Maia, paves the way for training more capable models and making those models cheaper for our customers," Altman added.

Elsewhere, the Microsoft Azure Cobalt 100 CPU will be optimized to deliver greater efficiency and performance in cloud-native offerings, being built on Arm architecture to provide maximum performance per watt across more general purpose workloads.

Microsoft says that having the ability to build its own custom silicon allows it to target certain qualities and ensure a high level of performance even on the most demanding workloads.

“Microsoft is building the infrastructure to support AI innovation, and we are reimagining every aspect of our datacenters to meet the needs of our customers,” said Scott Guthrie, executive vice president of Microsoft’s Cloud + AI Group. “At the scale we operate, it's important for us to optimize and integrate every layer of the infrastructure stack to maximize performance, diversify our supply chain and give customers infrastructure choice.”

More from TechRadar Pro

• Someone has used diamonds and a high-end Nvidia GPU to build perhaps the world's fastest graphics card ever — but you won't be able to buy it yet
• The first AI nation? A ship with 10,000 Nvidia H100 GPUs worth $500 million could become the first ever sovereign territory that relies entirely on artificial intelligence for its future
• These are the best AI writers available today

Microsoft and Oracle have entered a new multi-year partnership that will not only see cloud and AI companies face a greater threat, but also search.

The agreement will see Microsoft use its own Azure AI infrastructure along with Oracle Cloud Infrastructure (OCI) AI to make its search engine even more powerful, and Google may want to start worrying.

Redmond plans to use the two AI infrastructure services for the inferencing of AI models which are going on to power conversational searches in Microsoft Bing.

Bing’s appetite for AI infrastructure is growing

While the agreement between two rival companies might seem counterintuitive, Microsoft has nothing to worry about. Its 23% share in the cloud market in Q3 2023 (via Synergy) is leaps and bounds ahead of Oracle’s, which doesn’t even make the top five. However, Synergy noted that Oracle saw healthy year-on-year growth figures, and the extra support from Microsoft could see it make moves among tier-two cloud providers.

Microsoft’s global head of marketing for Search & AI, Divya Kumar, said: “Our collaboration with Oracle and use of Oracle Cloud Infrastructure along with our Microsoft Azure AI infrastructure, will expand access to customers and improve the speed of many of our search results.”

Karan Batta, SVP for OCI, added: “By furthering our collaboration with Microsoft, we are able to help bring new experiences to more people around the world.”

The joint effort takes advantage of Oracle Interconnect for Microsoft Azure, which enables interoperability between the rival companies’ cloud infrastructures.

Clearly, Microsoft has outgrown its own resources as it looks to divvy up some for itself and some for its customers, amid what Oracle is calling an “explosive growth of AI services.”

It’s unclear how much custom Microsoft will give Oracle, and how long the partnership will last. TechRadar Pro asked both companies to confirm – Oracle did not immediately respond. Microsoft told us that it had "nothing new to share."

More from TechRadar Pro

• Microsoft and Oracle reveal major cloud tie-up
• We’ve rounded up a list of the best cloud hosting providers
• Rather use existing services than build your own? Here are the best AI tools and best AI writers

The cloud-based identity and access management service that many of us have come to know as Azure Active Directory is becoming Microsoft Entra ID as the company announces a rebrand.

The move comes as part of Microsoft's efforts to simplify nomenclature, with what is now being known as Entra ID sitting alongside Entra Permissions Management and Entra Verified ID since Entra’s launch in May 2022.

Happily, for users, this means nothing more than becoming familiar with the product’s new naming.

Azure AD becomes Microsoft Entra ID

Microsoft confirmed the precise standalone license name changes: “Azure AD Free becomes Microsoft Entra ID Free. Azure AD Premium P1 becomes Microsoft Entra ID P1. Azure AD Premium P2 becomes Microsoft Entra ID P2. And our product for customer identities, Azure AD External Identities, becomes Microsoft Entra External ID.”

However, changes will be gradual, with SKU and service plan name changes not taking effect until October. Completion of the name change is expected to occur before the end of 2023.

Microsoft customers concerned about how the changes may affect them as the Entra sub-brand begins to take over some operations are being told not to worry:

“Capabilities and licensing plans, sign-in URLs, and APIs remain unchanged, and all existing deployments, configurations, and integrations will continue to work as before.”

At the same time, Redmond announced the availability of Microsoft Entra Internet Access and Microsoft Entra Private Access, both of which are designed to make securing access easier and more effective in an era when cyberattacks are becoming more prevalent.

Moreover, they join a growing suite of Entra-branded products designed to give admins control over identity and access management, and network access.

• Looking to move some things off-prem? Check out the best cloud hosting providers and best cloud storage

Unredacted court documents first reported on by The Information reveal that Microsoft’s Azure business generated $34 billion in revenue in the 12 months leading up to June 2022 - less than half (47%) of the $72 billion generated by its key rival AWS over the same period.

Microsoft has long been secretive about the true scale of its cloud business, instead leaving investors and analysts to merely guess how well it has performed. The court documents, which have now been removed, may push Redmond to now be somewhat more transparent about its revenue.

Instead, the company has historically packaged Azure figures together with Enterprise Services, Windows Server, and SQL Server figures, in a business operation it calls Intelligent Cloud.

Microsoft revenues

The court filing, which relates to yet another antitrust case in the US (Microsoft is the hot topic in the EU for its reportedly anticompetitive cloud business), indicates that there’s no evidence if Microsoft’s Azure business is even profitable at this point.

The year following June 2022 has been a massive one for the company on the whole, with significant generative AI developments that have seen Azure play host to the OpenAI Service and many other artificial intelligence operations. Whether that means a spike in revenue, or a bigger spike in costs, it remains to be seen just how profitable Azure actually is.

Regardless, with both AWS and Azure accounting for 30-40% each of the UK public cloud market in the UK in 2021 (via Ofcom), alarm bells are beginning to sound about Azure’s comparably lower revenue.

In terms of the cloud infrastructure services market, Synergy suggests that AWS accounts for 32%, with Azure trailing behind in second place with 23%. 

In response to a request to comment, Microsoft told TechRadar Pro: "Microsoft has nothing to share."

• Looking to move your data off-prem? Check out the best cloud storage providers and the best cloud backup tools

We list the best cloud computing provider, to make it simple and easy to manage your data and software online, with robust security and at an affordable rate.

Cloud services have revolutionized computing, not least through IaaS, PaaS, and especially SaaS, which have allowed businesses to develop virtualized IT infrastructure and deliver software through the cloud, independent of a user's operating system.

Even better still, businesses can also mix and match cloud services from different providers to ensure maximum efficiency and cost effectiveness. For larger businesses especially the economic benefits can be significant.

Because cloud services are run through software platforms and virtualized networks, it means that it's easy to access and analyze data for the purposes of analytics as well as for business intelligence purposes. The result is IT infrastructure management that allows for insights that would have previously been much more difficult to access.

Cloud computing services also offer the advantage of being scalable, which means not only can you access additional resources exactly as you need them, but you are also charged only for the services you use so there's no need to buy in extra hardware for additional redundancy.

This especially applies to when storing data, as online cloud storage can be treated as effectively limitless. Even though you might be using cloud databases for your structured data, you can also archive everything else into massive data lakes for additional processing using AI and machine learning for greater insights.

Altogether, cloud services offer unparalleled potential for improving business performance and increasing profits, and here we'll look at the best cloud computing service provider.

There are endless cloud computing services out there, and choosing the best one can be an uphill battle. It’s essential to select an ideal one based on your requirements. Otherwise, you might find it strenuous to build and maintain your platform.

There are many things to consider including cost, pricing, security, features, ease of use, etc. We tested various cloud computing vendors and selected the best ones for you. Each one has its advantages and disadvantages but, overall, are some of the best cloud platforms you can find.

We've featured the best business cloud storage.

The best cloud computing service of 2026 in full:

Best cloud computing provider for reliability

Amazon Web Services (AWS) is the world’s biggest cloud computing provider. It kickstarted the cloud revolution and has occupied the number one spot for a long time.

AWS has the largest data center network of all cloud providers, so it’s very reliable and offers solid performance. You can host your application in the region where your customers are closest. For instance, if you’re in the U.K. and observe that most of your customers come from the U.S., you can choose a U.S.-based data center to offer faster speed.

AWS is known for being user-friendly. It's built for quick and easy access, and you can monitor or modify your data just when you want. This cloud platform offers a lot of web application services, so you have a broad choice of features to adopt. You can choose the best server operating system depending on which programming languages power your app.

AWS is also known for being very safe and secure. The platform has had few noteworthy security incidents since its founding in 2006.

However, there are some drawbacks. A common customer complaint is that AWS’s billing system is confusing, leaving users prone to making mistakes that increase their costs. Another complaint is the high costs of dedicated support from AWS experts.

• ^ Back to the top

Best cloud computing provider for scalability

Azure is a cloud computing service offered by Microsoft, the American tech giant best known for its Windows operating system. It allows you to build and run your applications on the cloud with ease.

Azure has a large network of over 200 data centers spread across the globe. This network guarantees high availability and redundancy, with Azure offering a service level agreement (SLA) of 99.95% (roughly 4.38 hours of downtime per year).

Azure has invested a lot of resources into security. It offers built-in controls for detecting and thwarting cyber threats; it applies artificial intelligence and machine learning to aid this process. The company has a team of over 3,500 cybersecurity experts working to safeguard its cloud network.

Scalability shouldn't be a worry when using Azure. You can start with small computing resources and increase them as your business grows. The platform uses the pay-as-you-go method, making it cost-effective as you scale.

But a notable drawback of using Azure is its complex user interface. You’ll likely find it more difficult to navigate Azure than AWS and some other rival platforms.

Take a look at our Microsoft discount codes for the best Microsoft offers and savings.

• ^ Back to the top

Best cloud computing provider for Google

Google Cloud is an eponymous cloud computing service offered by Google, the tech company best known for its search engine. Google got into the cloud game later than most competitors but nonetheless has built a strong customer base.

Google Cloud dedicates a lot of resources to security. It uses artificial intelligence and machine learning systems to detect and counter cyber threats. Users have access to a host of cybersecurity features to keep sensitive data secure.

This platform is famous for its high performance. Ask many customers the best thing about choosing Google Cloud, and you’ll likely hear about performance. Google has a lot of hardware and software expertise that it leveraged to design servers and server management software that works as fast as possible.

Another good thing you’ll hear about Google Cloud is that the platform makes it easy for multiple users to collaborate within the same interface and access the same features. This platform also offers comprehensive support resources, including tutorials, how-to articles, an online community, and human support.

The main complaint that you’ll hear about Google Cloud is that it’s expensive, which is true. It costs more to use Google Cloud than rival platforms like AWS and Azure, but many people think the tradeoff for high performance is worth it.

• ^ Back to the top

Best cloud computing provider for enterprises

IBM Cloud is a cloud computing service offered by the eponymous tech giant IBM. The solution offers PaaS, IaaS, and SaaS.

As hardware and virtual-based servers are combined into one on demand cloud-platform, you have complete control. IBM refer to their hardware servers as ‘bare metal’. These provide clients with sole access to their entire server. This reduces the ‘noisy neighbor’ effect and greatly improves performance.

IBM Cloud is integrated and managed by a single system that can be controlled via web portal, API (Application Programming Interface) or mobile apps. IBM Cloud’s Bluemix developing solution has a wide range of cloud SaaS management tools.

IBM Cloud also offers full server customization. This means that everything that is included in the server is handpicked by you. This way you do not have to pay for features that you might never use.

• ^ Back to the top

Best distributed cloud computing provider

Oracle Cloud Infrastructure offers two main service provisions available: cloud architecture, and storage data.

Cloud architecture includes data management, cloud databases, and applications, while the Oracle Data Cloud is primarily for driving big data analytics for business intelligence insights. Oracle also provides a range of SaaS platforms such as HCM, EPM, SCM, and social media tools.

For those who do sign up, aside from a 30-day free trial, Oracle Cloud also offers a free service tier which includes unlimited access to two autonomous databases which come with Oracle Application Express (APEX) and Oracle SQL.

• ^ Back to the top

Best open source cloud computing OS

CloudLinux isn't so much a cloud computing provider, but rather a cloud platform you can build across your own servers. This means that if you'd prefer to have tight control over your cloud network rather than going with third-parties, you can host it yourself. While this presents a different set of challenges, it also offers a range of benefits, especially for those companies already heavily invested in their own IT infrastructure.

CloudLinux is a distro of specifically built for virtualized cloud operations to power shared hosting services, by allocating and limiting server resources to each tenant. Additionally, each tenant is partition from each other, in other to create a more secure environment. The result is a hosting platform that has built in redundancy, stability, and security, in order to provide a better hosting environment.

All cloud services require a degree of technical knowledge to work with, so for those with extensive expertise CloudLinux offers the opportunity to be far more in control of your own services and how they are configured.

• ^ Back to the top

Best cloud computing provider for Europe

OVHcloud is a French cloud computing provider focused on European customers. It is the largest cloud hosting provider in Europe and the third-largest globally based on physical servers.

OVHcloud gives you access to virtual private servers or dedicated servers. You can choose the type you want and the region you want it to be in. Your choice depends on the type of online platform you're hosting, as some servers are faster than others.

This platform has the largest data center network in Europe, so it’s an ideal choice if most of your customers are based in this region. Furthermore, OVHcloud is affordable compared to most rival platforms.

When testing OVHcloud, we observed that its customer support wasn’t as good as what you’d get on rival platforms. You need to pay a significant sum for dedicated human support, which puts it at a disadvantage.

• ^ Back to the top

Best cloud computing provider for cost-effectiveness

DigitalOcean is one of the most popular low-cost cloud providers. It doesn’t offer as much sophisticated features as you’d find on rival platforms like Azure, AWS, and Google Cloud. It focuses more on offering simple cloud computing features for an affordable price and does its job well.

DigitalOcean has 15 data centers worldwide, much smaller than the network of the likes of AWS and Azure. Its reliability and performance are good but not as good as what you’d get on these rival platforms. The attraction of DigitalOcean is trading lower costs for fewer features and performance.

It's pretty easy to spin up cloud servers on DigitalOcean to host your application. The platform is easy to navigate, so you shouldn’t face difficulties navigating it. You can choose different types of managed cloud databases, clusters, and APIs. You can also access complementary features such as security firewalls, load balances, and a DNS management system.

• ^ Back to the top

Best cloud computing provider FAQs

How to choose a cloud computing provider 

Cost

Cost is the number one consideration in selecting a provider. It’s essential to pick something that you can afford to avoid running into trouble in the future. Cloud services are unique such that they don’t usually have a straightforward pricing structure. What you’ll pay depends on the features that you select. The good thing is that cloud services offer their pricing details beforehand so that you can have a good estimate of what you’ll pay.

Security

You’ll be storing a lot of sensitive data on your cloud provider. Thus, it must have advanced security features to prevent data leaks or theft. Losing customer data can hamper your reputation and even bring about legal lawsuits. Similarly, losing sensitive business data can harm your competitive advantage, so your cloud platform must be secure.

Features

What features do the platform offer? Can you choose the type of servers you want? Can you spin up GPUs? Does it offer complementary services such as domain names and DNS management? Etc.

Ease of use

You need a platform that’s easy to understand and navigate. Otherwise, you can get frustrated or make costly mistakes by choosing features you didn’t intend to choose.

What are the essential features of a cloud provider? 

On-demand self-service

You should be able to spin up cloud instances and use them without external help. You should be able to monitor the servers you've rented and their allotted storage capacities. External help can always be a complementary feature, but the important thing is that every good cloud provider must offer a self-serve option.

Solid uptime

A cloud provider should have their servers up and running as faultlessly as possible. Of course, no platform is perfect, so you may experience occasional downtimes. However, every cloud provider should offer at least 99.9% uptime. Otherwise, you may lose money because of unnecessary server issues preventing your online platform from working.

Analytics

An ideal cloud computing platform should let users analyze what features they’ve used and give them a breakdown of how it affects their costs. This way, a user can easily identify which features are generating high costs and which ones aren’t, and use this information to optimize their spending.

Pay-As-You-Go

Pay-as-you-go is the best pricing structure because you only pay for the computing features you use. This way, you avoid paying for unnecessary things that you may not use, and you save significant sums in the long run.

How we tested the best cloud computing providers

We test by evaluating numerous factors. To start with, we look at the feature set, the range of tools available, and what range of business type and size the cloud computing service would be ideal for. We consider how easy the setup is, the simplicity of the interface, and whether there's sufficient documentation and tutorials for users to utilize necessary options optimally.

We assess how well the service integrates with other relevant services, such as security perimeters and user apps. We also looks at pricing models and flexibility, and lastly, we judge the quality of the customer service and the different pricing plans available.

Read more on how we test, rate, and review products on TechRadar.

Get in touch

• Want to find out about commercial or marketing opportunities? Click here
• Out of date info, errors, complaints or broken links? Give us a nudge
• Got a suggestion for a product or service provider? Message us directly

• You've reached the end of the page. Jump back up to the top ^

Microsoft has announced its Azure OpenAI Service is now available to the US government in an even more secure format, with specific compliance promised for regulatory standards for classification and security.

With Azure OpenAI Service for government, customers (which currently include United States federal, state, and local government and their partners) can get access to OpenAI’s GPT large models - those behind the popular ChatGPT AI writer.

Microsoft hopes that moving those models from the commercial cloud to a new and secure architecture will enable the government to benefit from the time-saving tech that’s stolen many of the headlines in recent months.

Azure OpenAI Service for government is now live

The service’s REST APIs give the government access to language and multimodal models including GPT-3, GPT-4, as well as Embeddings. 

The company expects government agencies to leverage these existing models to develop their own AI-enabled applications to help improve efficiency across their operations.

With Azure OpenAI Service for government comes the usual list of benefits and use cases that can be expected, including content generation, summarization, semantic search optimization, and code generation and rectification. 

In a bid to convince government bodies that its services are secure, Microsoft sheds some light on the 250,000 km of fiber optic and undersea cables that make up its global network backbone. Government data promises never to leave this and enter the public Internet.

In the announcement, Microsoft CTO Strategic Missions and Technologies, Bill Chappell, explained: “Only the queries submitted to the Azure OpenAI Service transit into the Azure OpenAI model in the commercial environment through an encrypted network and do not remain in the commercial environment.”

Furthermore, Chappell confirmed that government data is not used for training OpenAI’s models in the same way that regular consumer information is used.

How the US government plans to use generative AI to its benefit remains to be seen, but its approach is likely to be cautious at first as it gets to grips with integrating artificial intelligence with the sensitive data it handles.

• Check out our roundup of the best cloud storage companies

A Microsoft Azure DevOps outage in the South Brazil Region, which lasted over 10 hours, was caused thanks to a typo in the code that saw 17 production databases deleted.

Having apologized to impacted customers for the outage, Microsoft has now issued a full post-mortem, sharing details about the investigation that took place from when the outage was first noticed at 12:10 UTC on May 24, until its remedy at 22:31 UTC on the same day.

Microsoft principal software engineering manager Eric Mattingly shared details of the code base upgrade which formed part of Sprint 222. Inside the pull request was a hidden typo bug in the snapshot deletion job, which ended up deleting the Azure SQL Server rather than the individual Azure SQL Database.

Coding error

Mattingly explained: “when the job deleted the Azure SQL Server, it also deleted all seventeen production databases for the scale unit,” confirming that no data had been lost during the accidental process.

The outage was detected within 20 minutes, at which point the company’s on-call engineers got to work, however according to the event log the root cause was identified at 16:04, almost four hours after the outage had begun.

Microsoft blamed the over ten-hour fix time on the fact that customers themselves are unable to restore Azure SQL Servers, as well as backup redundancy complications and a “complex set of issues with [its] web servers.”

Having learned from its mistake, Microsoft has no promised to roll out Azure Resource Manager Locks to its key resources, in an effort to prevent future accidental deletion. 

Despite a same-day fix, customers in the region were left without access to some services for several hours, emphasizing how easy it is for things to go wrong and the importance of having backup plans to reduce reliance on single service providers, including cloud storage and other off-prem infrastructure.

• Looking for an alternative? Check out the best CDN providers

Microsoft’s Azure Active Directory (Azure AD), the company’s go-to cloud-based identity and access management service (IAM), carries a severe flaw that enabled threat actors to install backdoors.

This is according to extensive research from the Secureworks Counter Threat Unit (CTU), which says the issue could also let hackers modify access rights to bypass multi-factor authentication and block admin access without proper logging, and gather information on policy configurations to enable future attacks.

Azure AD supports multiple authentication methods, while the premium version also supports Conditional Access Policies (CAPs) that grant, or block access, based on different criteria, such as device compliance or user location. The IAM service is the one storing these settings, allowing CAPs to be modified either via the portal, PowerShell, or API calls. 

One API

The researchers set out to see which APIs allow CAP settings editing, and found three. 

One of the three, called AADGraph, was the only one allowing users to modify all CAP settings, including the metadata. This, the researchers say, allows admins to tamper with things such as creation and modification timestamps, and given that modifications made using AADGraph weren’t being properly logged, the integrity and non-repudiation of Azure AD policies were thus at risk.

The researchers shared their findings with Microsoft in late May 2022, which confirmed the findings a month later but stated that this was not a bug, but a feature. However, a year later, Microsoft notified CTU researchers that it plans on making changes that will improve audit logs and restrict CAP updates via AADGraph.

Secureworks also stresses that Microsoft’s been trying to deprecate the AADGraph API “for years”. At the moment, the retirement is scheduled for June 30, 2023. Microsoft has removed public AADGraph API documentation.

• These are the best firewalls to keep your business protected

US and other non-EU cloud providers will have to forfeit some of their control to EU-based partners if they are to maintain their position in the European market, reports have claimed.

According to a new draft document seen by Reuters, the proposed change will see non-EU-based cloud service providers having to enter a joint venture with EU-based bodies in order to obtain an EU cybersecurity label, which allows them to handle sensitive data.

Should it be approved, it will affect the three most prominent cloud providers that occupy between two-thirds and three-quarters of the entire market for both the personal and business sectors: Amazon, Microsoft, and Google.

US cloud in the EU

As well as having to enter a partnership with an EU company, staff chosen to handle sensitive data will have to be located within one of the Union’s member states and have undergone specific screening.

A ‘high+’ level appears to have been added to the ‘basic,’ ‘substantial,’ and ‘high’ tiers that form categories of The Cybersecurity Act, whereby ‘high+’ would require an EU company to have total control over the cloud service “to mitigate the risk of non-EU interfering powers undermining EU regulations, norms and values.”

‘High’ and ‘high+’ would also be subject to data localization measures within the EU.

Despite the apparent effort to protect the EU as a whole, one industry source suggested that each country would have the discretion to impose requirements as it sees fit (via Reuters).

It is believed that the European Commission may implement a final version of the rules once they have been agreed by the countries that make up the European Union, following discussions at the upcoming ENISA Cybersecurity Certification Conference.

• Need to check out alternatives? These are the best cloud backup tools around

Microsoft has announced the general availability of two new Azure instances with remote storage performance and memory-intensive business-critical applications in mind.

The new Ebsv5 and Ebdsv5 models promise up to three times higher remote storage IOPS and up to six times higher remote storage throughput performance than the Ev4 that succeeded them.

With the higher performance, Microsoft’s suggestion is that customers may be able to reduce the number of VMs they require, saving some cash.

Ebsv5 and Ebdsv5 NVMe VM series

Both the Ebsv5 and Ebdsv5 NVMe virtual machines will offer up to 260,000 IOPS and 8,000 MBps of remote disk storage throughput. They can be configured up to 672 GB of memory and the Ebdsv5 can support up to 3,800 GB of storage provided by SSDs.

The higher figures are possible thanks to the introduction of new E96 and E112i vCPUs, with the lowest Standard_E2bs_v5 VM capped at 16GB memory, 5500 IOPS, and 156 MBps of remote disk storage throughput.

It’s worth pointing out that the Ebsv5 family doesn’t include local storage, unlike the Ebdsv5, though both are based on the 3rd Gen Intel Xeon Platinum 8370C chips.

Microsoft explained in its announcement: “In today's data-driven business environment, rapid processing, analysis, and extracting actionable insights from the increasing volume of data collected is critical for companies to maintain a competitive edge,” citing online transaction processing as an example use case.

The new NVMe-enabled VMs will be available in the following 13 regions with more set to follow, and will join the Ebsv5 and Ebdsv5 SCSI family that launched a year ago:

USNorth, Southeast Asia, West Europe, Australia East, North Europe, West US 3, UK South, Sweden Central, East US, Central US, West US 2, East US 2, Southcentral US.

• After somewhere to store your data? How about the best cloud storage or best cloud backup tools?

Microsoft has generally released its multi-cloud security benchmark (MCSB) for enterprise, but we can’t decide whether it’s as much an advertisement for its Azure cloud storage or a boast that the security of competing services just isn’t good enough.

Per The Register, the confusion comes from the fact that, while Microsoft is still providing advice on how to secure enterprise instances of Azure with MCSB, it’s now also doing the same for AWS, and will do the same for Google Cloud later in 2023.

That in itself could be considered reasonable, and even well-intentioned - except for the fact that the new MCSB v1 also offers automated monitoring features (172 separate checks, if you’re counting) for AWS at this time.

Microsoft and multicloud

The Register reported Jim Cheng, a senior software engineer at Microsoft, as saying that, because multicloud configurations are on the rise, it’s important for the tech giant to be “looking outward”.

"[Aggregating security management],” he claimed, “often requires security teams to repeat the same implementation, monitoring, and assessments across different cloud environments and often for different compliance standards. This creates unnecessary overhead, cost, and effort."

From this, it’s clear that Microsoft is pitching the MCSB as a gesture of goodwill, but it’s worth noting that it’s keen to bake it into Windows 11 as a value offering. The MCSB is enabled and governed in Microsoft Defender for Cloud, which the company are positioning as a centralized security dashboard.

That’s important, given that, once Google Cloud monitoring comes to the MCSB, organizations will be able to monitor three major cloud providers (comprising 66% of the market) all from one Microsoft-exclusive application.

•  Here’s our list of the best endpoint protection right now

Since ChatGPT launched in November 2022, individuals and businesses have been using it to summarize text, create new content, think up new ideas, and even write code.

As the service began to amass millions of active monthly users, rumors started circulating about Microsoft’s involvement with the company behind ChatGPT, OpenAI. Since then, Microsoft is reported to have invested billions in the company and artificial intelligence has been making its way into numerous products and services.

This partnership has now taken a significant step forward, as Microsoft will be integrating ChatGPT directly into its Azure OpenAI Service, extending its reach to millions of customers through the large number of companies that use Azure.

ChatGPT in Azure

It’s important to note that ChatGPT is currently in preview in Azure OpenAI Service, and full general availability may be some time off yet with future adjustments to services and pricing both possible and indeed likely. For now, though, it’s priced at $0.002 per 1,000 tokens.

Microsoft hopes that its Azure customers will use the AI chatbot to integrate into their own experiences and applications, such as enhancing existing bots, summarizing calls and meetings, creating new marketing material, and automating claims processes.

For anybody still unsure how AI can benefit their business, Microsoft cites some third parties in a blog post who have experienced enhancements, as well as some of its own services including GitHub Copilot, Teams Premium, Viva Sales, and Bing.

Overall, Microsoft’s hope with Azure OpenAI Service is to democratize artificial intelligence and get it into the hands of as many customers as possible, however some argue that the global rollout has been too slow with too much of an emphasis on certain parts of the world, likely due to how difficult it is to deliver low-latency AI on a large scale. 

• These are the best cloud hosting providers and best cloud storage solutions available for you right now

Microsoft has transitioned its confidential containers on Azure Container Instances (ACI) from limited preview to public preview as full general availability moves one step closer.

A Microsoft blog post explains how the service makes use of Secure Encrypted Virtualization and Secure Nested Paging (SEV-SNP) in AMD’s latest chips designed for servers.

ACI Product Manager Peter Pogorski explained how Azure customers are “increasingly turning to cloud-native, container-based applications to support their workloads”, but they continue to demand high levels of data protection.

Azure confidential containers

“This serverless platform allows for running Linux containers within a hardware-based and attested Trusted Execution Environment (TEE), providing the simplicity of a serverless container platform with the enhanced security of confidential computing," Pogorski summarized.

Data-in-use can be protected in confidential containers by being processed in encrypted memory, which Microsoft has AMD to thank for with its EPYC processors.

Confidential containers are designed to run with verifiable initialization policies meaning that Azure’s customers can also ensure that executed code is trusted and verified, helping to eliminate unintentional data leaks.

For collaboration, guests can also review the attestation report to make sure that they’re happy with an application running in a container group before they commit to sharing sensitive information. 

ACI currently supports things like batch processing, data processing pipelines, and continuous integration, but Microsoft reckons confidential containers will open up its services to even more scenarios.

In its limited preview announcement, Azure Confidential Computing Product Manager Amar Gowda said: 

“We are excited to bring confidential serverless offerings with full lift & shift container support while continuing to innovate in this fast-emerging confidential computing and cloud native space.”

While there has been no official indication of full general availability, the fast-paced development suggests that confidential computing is key for many businesses as they seek to optimize their IT.

• Check out our picks for the best cloud hosting providers and best web hosting providers

Microsoft has announced details of a new Azure platform that it says will see cloud and edge come closer together through the power of 5G.

Speaking at its MWC 2023 launch, Microsoft says it’s time to move away from legacy systems as 5G development continues to accelerate, citing improved bandwidth, reliability, and reduced latency as some of the key benefits of 5G connections.

The latest iteration of Azure for Operators will allow companies to modernize their networks, resulting in improved cost efficiency and the rollout of new services.

Microsoft Azure for Operators

For Azure for Operators, Microsoft has moved the engineering and program organization from AT&T’s Network Cloud 2.7 Software technology (which it acquired in June 2021) into Azure for Operators. 

Microsoft claims that Azure Operator Distributed Services combines the best of AT&T’s and Azure’s offerings, such as security, monitoring, analytics, AI, and machine learning.

"Imagine the benefits to communities and organizations that have access to improved bandwidth, reliability, and reduced latency, while leveraging the rich capabilities of cloud-to-edge technology without compromise to security, critical services, or key workloads," noted Jason Zander, Microsoft EVP, Strategic Missions and Technologies.

"With the most complete offerings for the telecommunications industry, Microsoft is the ideal cloud provider to help operators with their digital transformation journey and enable them to deliver these innovative services to their consumer, enterprise, and public sector customers."

“We are pleased with Microsoft’s plan to evolve Network Cloud and integrate it with Azure technologies to create hybrid telco-grade Azure Operator Distributed Services," noted AT&T Executive VP and CTO of Network Services, Andre Fuetsch.

"This will enable AT&T and other operators to host Network Functions on clouds spanning telco premises and public cloud and will help us realize the many benefits of the cloud-native approach and Azure innovation including additional speed, resiliency, security, cost, and operational improvements.”

At the same time, Microsoft has announced the private preview of the Azure Operator 5G Core and public preview of the Azure Private 5G Core, designed with the secure deployment of private 4G and 5G networks in mind. With Industry 4.0 and IoT at their hearts, use cases could include smart cities and autonomous vehicles. 

• Need somewhere to store your data now? These are the best cloud backup tools

Microsoft has announced the general availability of its Azure OpenAI Service, which incorporates AI writer tools like GPT-3.5, Codex, and DALL-E 2, in a move that it hopes will democratize artificial intelligence technologies and make them more readily available to more customers. 

The tech giant debuted Azure OpenAI Service back in November 2021, offering up a combination of some of the most powerful AI tools we have come to know with the characteristics typical of Azure, which it says are security, reliability, compliance, and data privacy.

In the space of just a couple of months, the Azure platform claims to have freed up workers’ time by generating things like support ticket summaries and generating more relevant content for online shoppers. 

Microsoft Azure OpenAI Service

Microsoft invested $1 billion in OpenAI - the company behind the popular GPT model - and is reported to have invested a further $2 billion since then according to The New York Times. 

Fast forward to 2023 and the company is said to be preparing a large $10 billion investment in the company and integrating its AI into more Microsoft services, including its Bing search engine.

“We have taken an iterative approach to large models, working closely with our partner OpenAI and our customers to carefully assess use cases, learn, and address potential risks," Microsoft AI Platform corporate VP, Eric Boyd, explained in a blog post.

Some measures to protect users include the requirement for developers to apply for access and further content filters.

While keen users can apply for Azure OpenAI Service now, The Register has highlighted its distinct lack of general availability, rendering it only available in four US-based Azure regions and the West Europe region, which it says is down to the large amount of effort required to make low-latency AI available globally.

• We’ve compiled a list of the best live chat software

Microsoft is set to purchase an “approximate” 4% stake in the London Stock Exchange Group (LSEG) as part of a new strategic partnership that will se the LSEG’s data infrastructure migrate to Microsoft’s cloud hosting platform Azure.

The financial data provider also claims that the Workspace platform will become “interoperable” with the video conferencing platform Teams and the office software suite 365.

In a press release, the LSEG revealed that the company has committed to a “minimum cloud-related spend” of $2.8 billion (£2.3 billion), with additional spend conditional on how Microsoft’s involvement drives demand for LSEG’s financial analytics platform Workspace and “other related services”.

Microsoft and the LSEG’s partnership

It’s not all about product migration, though. The LSEG hopes to develop “new products and services for data and analytics” alongside Microsoft, following the former’s acquisition of Refinitiv, another financial data provider, in February 2021.

“This strategic partnership is a significant milestone on the LSEG’s journey towards becoming the leading global financial markets infrastructure and data business, and will transform the experience for our customers, said David Schwimmer, the LSEG’s CEO.

Satya Nadella, Chairman and CEO at Microsoft said that “our partnership will bring together the industry leadership of the London Stock Exchange Group with the trust and breadth of the Microsoft Cloud, spanning Azure, AI, and Teams to build next-generation services that will empower our customers to generate business insights [and] automate complex and time-consuming processes.”

So, both companies are claiming that the partnership will change the way financial sector customers conduct business, but it’s also, shock horror, about money. 

“Bringing together our leading data sets, analytics, and global customer base with Microsoft’s comprehensive and trusted cloud services and global reach creates attractive revenue growth opportunities for both companies,” Schwimmer continued.

The LSEG estimates that the costs of the partnership between 2023-2025 will be between £250-£350 million. Despite this, and the mammoth minimum spend, it also believes that its revenue will see a “meaningful boost” over time as new jointly developed analytics products become available for its customers.

The partnership with the LSEG will continue to frustrate Microsoft’s competitors in the video conferencing and collaboration space as it continues to throw money at securing a monopoly.

In November 2022, Slack threatened an antitrust lawsuit in the European Union over Microsoft Teams’ sheer dominance, and it’s certainly likely that the LSEG partnership will drive even more users in the financial sector over to the service. 

•  Here’s our list of the best productivity tools right now

NVIDIA and Microsoft are collaborating on a new cloud-based AI-focused supercomputer, which they claim will be "one of the most powerful in the world" when complete.

The new machine will leverage the supercomputing infrastructure of Microsoft Azure combined with NVIDIA GPUs, networking, and AI software. It's set to contain ND- and NC-series virtual machines specifically designed for AI distributed training and inference. 

The companies claim the project represents the first public cloud to incorporate NVIDIA’s full AI stack and will add tens of thousands of NVIDIA A100 and H100 GPUs, NVIDIA Quantum-2 400Gb/s InfiniBand networking, and the NVIDIA AI Enterprise software suite to its platform.

How will it be used?

The firms said the new machine will be used to help enterprises train, deploy and scale AI, including large models.

NVIDIA is also set to utilize Azure’s scalable virtual machine instances to research and further advances in generative AI. 

This is an emerging area of AI in which foundational models like Megatron Turing NLG 530B provide the basis for unsupervised, self-learning algorithms to create new text, code, digital images, video or audio.

The companies will also collaborate to optimize Microsoft’s DeepSpeed deep optimization software and NVIDIA’s full stack of AI workflows and software development kits, optimized for Azure, will be made available to Azure enterprise customers.

“AI technology advances as well as industry adoption are accelerating. The breakthrough of foundation models has triggered a tidal wave of research, fostered new startups and enabled new enterprise applications,” said Manuvir Das, vice president of enterprise computing at NVIDIA.

It's not just Microsoft that is looking towards Nvidia to power its latest AI innovations.

Oracle and Nvidia announced a collaboration at Oracle Cloud World 2022. It will see tens of thousands of Nvidia GPUs, such as the A100 and upcoming H100, supporting Oracle Cloud Infrastructure (OCI).

• Interested in hosting more of your AI projects in the cloud? Check out our guide to the best cloud hosting providers.

WP Engine has launched its managed WordPress platform on Microsoft Azure for companies who prefer Microsoft.

Available now, businesses in the US and the Netherlands can choose Azure to build and scale WordPress websites with WP Engine.

WP Engine says it plans to expand into other regions throughout 2023 for a more global reach on the cloud hosting platform.

WP Engine on Microsoft Azure 

The new solution uses Kubernetes for containers, which allows engineers to port the platform to AKS without making core modifications to code. 

“Working with Microsoft on an Azure-based offering helps us deliver on our promise to be the world’s most trusted WordPress platform for businesses around the world,” said WP Engine Chairwoman and CEO Heather Brunner. 

“We’re proud to collaborate with Microsoft in powering their Microsoft Stories site, now 7X faster with WP Engine powered by Azure.” 

WP Engine says that the offering combines the agility of its WordPress hosting platform with the global reach of Microsoft Azure. It also confirmed that the solution offers “certified” security for enterprise and SMB businesses.

“WP Engine on Azure paves the way for Microsoft partners and customers to publish their own WordPress sites at scale,” said Sean McKenna, Director of Product, Azure Kubernetes Service at Microsoft. 

“The WP Engine platform integration with Azure Kubernetes Service (AKS) creates a powerful software solution that natively supports dynamic websites built on WordPress.” 

WP Engine provides managed WordPress hosting, enterprise WordPress, headless WordPress, Flywheel, Local, Advanced Custom Fields, and Genesis.

“Wrapping WordPress components in containers makes it easier for teams to use, manage and deploy applications anywhere,” said Ramadass Prabhakar, SVP and chief technology officer at WP Engine. 

“This modern design approach is an example of how containers can be leveraged to build a portable architecture.”

WP Engine began working with Microsoft in 2018 when Microsoft chose the WordPress platform to power its Microsoft Stories website.

In 2021, WP Engine and Microsoft Azure collaborated again as an increase in content demand and the streaming of in-house web processes led both companies to fully integrate its platforms.

Great news for WordPress users in the US and Holland but for those outside of those areas, you can check out our best WordPress hosting buying guide for other options.

• Check out our list of the best web hosting services on the market

Microsoft has made the latest iteration of its Azure Premium range of solid state drive (SSD) cloud storage available for general sale.

The company announced the release of the latest Azure Premium SSD v2 storage at its Ignite event, following an invite-only preview release in May 2022.

The new range of high-capacity drives for enterprises is designed to offer the lowest latency for “performance-critical workloads”, while also allowing for storage expansion up to 64 terabytes (TiB). 

Microsoft Azure Premium v2 improvements

Microsoft claims to have achieved “sub-millisecond” latency with the new drives, offering 80,000 input/output operations per second (IOPS) - a 60% improvement over Azure Premium V1. 

Data transfers should also be lightning fast, with a guaranteed baseline throughput of 125 megabytes per second (MBPS) across all drives under the new specification, and a peak throughput of 1,200 MBPS - an overall improvement of 33% over V1.

The increases to storage capacity may also be of interest to Azure Premium SSD v1 users who feel limited by the maximum 32 TiB available. 

Microsoft claims that the all-round increases to performance in the new drive range will suit a number of production workloads and purposes, from database software like SQL Server and Oracle, to analytics, virtual machines (VMs), and gaming.

The company is even looking to court enterprise customers who may not immediately need the benefits of Azure Premium V2, promising that companies may provision storage capacity, IOPS, and throughput “based on their workload requirements”.

This could make the new set of drives appeal even to smaller businesses who may go on to see steady growth, while simultaneously helping large companies cut their cloud costs during a recession.

•  Here’s our list of the best cloud hosting providers right now 

Microsoft is now publicly previewing a single sign-on experience for Microsoft Azure Virtual Desktop (AVD), enabling passwordless authentication for users.

Users will now be able to use Windows Hello and security devices such as FIDO2 keys to sign in, as well as third-party identity management services that integrate with Azure AVD.

In addition, users will be able to use passwordless authentication to sign both into the host computer using Azure ADV, as well as inside the session when using the Windows client.

How can I get started?

This new functionality is currently only available for Windows 10, Windows 11, and Windows Server 2022 session hosts, and only once you’ve installed the September Cumulative Update Preview.

You can follow the instructions from Microsoft on how to configure single sign-on, which will guide you in enabling the new authentication protocol. And if you’d like to get a quick overview of how the new functionality works as a whole, you can check out this intro video on Azure Academy.

In particular, if you’d like to use Windows Hello and security keys inside the session, Microsoft has also supplied instructions for in-session passwordless authentication explaining how to use the new WebAuthn redirection functionality.

If you are interested in trying out a virtual desktop, Microsoft isn’t the only game in town for Azure users.

Citrix has announced that its Desktop-as-a-Service (DaaS) offering is now available via Microsoft’s Azure Marketplace, which will reportedly simplify the delivery of Windows and Linux virtual desktops and applications to Microsoft customers.

Oracle VirtualBox users can also now run Windows 11, after the database giant addressed a known issue that caused the virtual machine to crash when it ran the operating system.

• Want to reduce the time you spend on passwords? Check out our guide to the best password managers

Alibaba Cloud is set to invest $1 billion over the next three years to support its partners' "technology innovation and market expansion".

The investment from the Chinese tech giant's cloud computing division will consist of a mix of both financial and non-financial incentives, including "funding, rebates, and go-to-market initiatives".

In addition, Alibaba Cloud is also set to launch a "regional accelerator" program to provide its partners operating in different markets with a localized business collaboration model.

So what's the bigger picture?

Alibaba already controls a pretty large proportion of the cloud storage market, according to statistics from Gartner, it trailed only Microsoft and Amazon with its 9.5% share of the market in 2021. 

The division already has around 11,000 partners worldwide, including Salesforce, VMware, Fortinet, IBM, and Neo4j.

Alibaba Cloud was already active in terms of looking to expand its international footprint.

Under its "Global Delivery and Service Program", it has unveiled three Customer Service Centers in 2022, located in Malaysia, Portugal, and Mexico, to support international customers during cloud adoption.

But unfortunately for investors, Alibaba's expansion outside of its core marketplace is seeing regulatory scrutiny.

In 2022, the US government was reported as reviewing Alibaba Cloud to determine whether or not it poses a risk to national security.

According to reporting by Reuters, the Biden administration's probe looked to ascertain how the company handled the data of US clients and whether the Chinese government could gain access to US intellectual property.

But it's not just the cloud computing market that the Hangzhou, China-headquartered company has its sights on taking over.

Alibaba has spun out a subsidiary, called Lingyang Intelligent Service Company, which it says will offer “enterprise digital intelligence services”.

The new business unit could potentially compete with the likes of Microsoft, Oracle, and SAP in the business intelligence space.

• Interested in trying out Alibaba Cloud? See if it features on our list of the best cloud hosting services 

Microsoft’s plans to extend its cloud operations into space are well underway, with the company having recently announced some new progression on its extra-terrestrial mission.

The latest announcement includes a preview of Azure Orbital Cloud Access, which the company says serves as a “step toward the future of integrated 5G and satellite communications”.

The low-latency cloud access, which works in conjunction with SpaceX’s Starlink, is designed to be available from “anywhere on the planet”, which will prove especially beneficial to companies outside of major cities where standard cellular signal is limited.

Microsoft Azure Orbital Cloud

As part of the announcement, Microsoft revealed the general availability of Azure Orbital Ground Station, which is a fully-managed ground station-as-a-service.

“We are bringing together a deep integration of ground station partner networks to enable our customers’ data delivery to an Azure region of choice at zero cost, thus reducing their total operational costs and ensuring data is available in the customer’s Azure tenant for further processing," said Microsoft in a statement.

The ground station became available on September 14, 2022, to “all satellite operators” including Pixxel, Muon Space, and Loft Orbital.

Further announcements included that SES, the company's Medium Earth Orbit (MEO) network partner for Microsoft Azure orbital, will be working with Microsoft to create a virtualized satellite communications ground network, which will apparently be a world first. Software-defined hubs, customer edge terminals, new virtual network functions, and edge cloud applications will be some of the key drivers behind the expansion of the joint venture.

Microsoft isn’t the only company with intentions to expand its cloud-based operations beyond the physical clouds. Segment leader AWS has its in-house Aerospace and Satellite program, while Google follows behind having reached a deal with SpaceX last year, according to an article from The Register.

For now, the Azure Orbital Cloud Access preview is only available to Azure Government customers, but if nothing else it provides an insight into what could reach consumer level in the coming years.

• These are the best dedicated server hosting providers for your business

Oracle's cloud-based database service MySQL HeatWave is now available to host on Amazon Web Services (AWS), the company has confirmed.

Launched in December 2020, HeatWave is designed to enable customers to run analytics on data that is stored in MySQL databases, without the need for ETL (Extract, Transform, Load).

Though Oracle's database service will be hosted on AWS, it is not delivered in partnership with Amazon and is directly competitive with the latter's own database services such as Amazon Aurora, Amazon RDS, or Amazon Redshift.

What does this mean?

Oracle claims the move will mean that users will be able to run transaction processing, analytics, and machine learning workloads in one service, without "requiring time-consuming ETL duplication" between separate databases

Oracle used the example of using Amazon Aurora for transaction processing, Redshift or Snowflake on AWS for analytics, and SageMaker for machine learning.

The new product apparently boasts delivers price performance that is 7X better than Amazon Redshift, 10X better than Snowflake, 12X better than Google BigQuery, and 4X better than Azure Synapse on the 4TB TPC-H* benchmark, if Oracle's claims are to be believed. 

The new product also brings new security features to the table, which apparently include server-side data masking and de-identification, asymmetric data encryption, and a database firewall

In addition, Oracle also that the service will also soon be available Microsoft Azure, as well as Oracle's own Oracle Cloud Infrastructure (OCI)

Oracle may be participating in activities that could threaten AWS's cloud dominance.

Microsoft is reportedly rallying other big players in the space, such as Google Cloud and Oracle in preparation to lobby the US government to ensure that large-scale cloud computing contracts are spread out between different vendors, what is known as a multicloud approach. 

"Many of our MySQL HeatWave customers migrated from AWS. Others wish to continue running parts of their application on AWS," said Edward Screven, Oracle chief corporate architect. "Those customers face serious challenges including exorbitant data egress fees charged by AWS and higher latency when accessing a database service running in Oracle's cloud".

He added: "We are addressing these issues while delivering outstanding performance and price performance across transaction, analytics, and machine learning compared to other database cloud providers."

The news comes as Oracle has posted strong corporate results in the quarter ending August 2022, Oracle's revenue grew 18% year-on-year to $11.4 billion, while its net income rose 14% to $1.5 billion according to its latest report.

• Also check out our guide to the best SMB servers

Microsoft Azure virtual machines (VMs) running the Linux distro Ubuntu 18.04, have been taken offline by a serious outage.

According to a statement from Microsoft, the issues began after Ubuntu 18.04 (bionic) VM users installed an update called "systemd version 237-3ubuntu10.54".

After installing the update, the impacted users began experiencing domain name system (DNS) errors.

Will this affect me?

Microsoft says the issue, which now appears to have been fixed was confined to this single Ubuntu version, however there was additional downstream impact on Azure services that leverage Ubuntu 18.04 (bionic) VMs including Azure Container Apps, Azure Database for PostgreSQL Azure Kubernetes Service, and Azure VMware Solution.

Ubuntu parent company Canonical did acknowledge the bug contained within the update, suggesting several fixes, however, it didn't mention the resulting Azure outagea. 

Dr. Nestori Syynimaa, a Senior Principal Security Researcher at Secureworks commented on the incident in a Twitter post, saying: "These things happen all the time and will happen as long people are using technology." 

He added: "Not on this scale that often though".

Outages on Azure can be extremely disruptive for organizations, as they form an important part of the critical operations of many organizations worldwide.

Taking steps to remedy this, Microsoft has announced that it will extend its “outage mode” in Azure for both web-based and desktop applications.

This means that if Azure Active Directory fails, the service will transition to “outage mode” where it can check requests and provide tokens, ensuring business continuity.

• Interested in an alternative to Microsoft Azure? Checkout our guide to the best servers.

Microsoft is set to launch Azure Virtual Machines (VMs) featuring the Ampere Altra Arm-based processor. 

Ampere is an American fabless semiconductor company based in Santa Clara, California, which launched in 2018 in a bid to take a bite out of the data center hardware market which had up until that point been dominated by Intel and AMD. 

Though hardware from UK-based Arm has long been omnipresent in the smartphone and consumer electronics space, x86 processors such as those provided by Intel and AMD have generally held sway in the cloud computing market.

What can these machines do?

The new Azure Arm-based virtual machines available include the Dpsv5 series, which offers up to 64 vCPUs and 4GBs of memory per vCPU up to 208 GBs.

In addition, customers will also be able to access the Dplsv5 series, which offers up to 64 vCPUs and 2GBs of memory per vCPU up to 128 GBs, and the Epsv5 series, which offers up to 32 vCPUs and 8GBs of memory per vCPU up to 208 GBs.

All the new virtual machine sizes will support up to 40 Gbps of networking bandwidth and support attached standard SSDs, standard HDDs, premium SSDs, and ultra disk storage.

The Dpdv5, Dpldv5, and Epdv5 virtual machine series also include fast local-SSD storage according to Microsoft.

But it's not just Microsoft that is looking to incorporate Arm-based VMs into their lineups.

The new Arm-based Tau T2A chips will join Google's existing line of Tau VMs, which were launched in June 2021.

Google says the new chips will be appropriate for scale-out workloads including web servers, containerized microservices, data-logging processing, media transcoding, and Java applications.

The new virtual machines will be generally available on September 1, and customers can already launch them in 10 Azure regions and in multiple availability zones around the world.

Prices will vary by region, and if you can't wait to get started, head here to start on Azure Arm-based Virtual Machines and AKS containers immediately.

• Interested in running VMs? Checkout our guide to best cloud hosting

Wider economic pessimism may finally be impacting the cloud computing market after revenue growth at all the major cloud providers slowed down in Q2 2022.

Amazon Web Services (AWS) posted sales of $19.7 billion for its second quarter, up 33% year-over-year, but with a 37% growth rate that was lower than the previous quarter.

Google Cloud brought in $6.3 billion of revenue in Q2 2022, a 35% year-on-year increase compared to Q2 2021’s $4.6 billion revenue, but much lower than the 44% jump from Q1 2021’s $4 billion to Q1 2022’s $5.8 billion.

Why the slowdown?

The Register pointed the finger at a mixture of supply chain concerns, increased spending on infrastructure, and increased electricity bills. 

Small data centers in the UK and Ireland have seen their margins significantly impacted as a result of the energy crisis, with some facilities seeing energy bills increase by as much as 50% over the last three years according to research by Aggreko.

Inability to keep up with demand  also could be a factor 

The Information reported that two dozen Azure data centers internationally are operating with limited capacity, citing two Microsoft managers with knowledge of the issue.

What are cloud firms doing?

It seems cloud firms are doing everything they can to keep costs low and profits high.

In 2022, AWS, Google, and Microsoft all added another year to the lifecycle of their cloud infrastructures, which could lead to significant savings across the board.

In Microsoft’s case alone the move could save $3.7 billion in its 2023 financial year alone, according to its own estimates.

But despite the possibility of slower growth on the horizon, the market as a whole for cloud hosting and cloud storage looks immensely healthy.

A recent Gartner report found the cloud market grew by 41.4% in 2021 to total $90.9 billion.

• Want to move away from cloud computing? Checkout out our guide to the best dedicated server hosting

Microsoft Azure is set to become the official cloud partner of the top Unity game engine.

Launched in 2005, Unity is a cross-platform game engine popular among indie game developers, and currently supports 19 different platforms including iOS, Android, and all the major current generation consoles.

It's been used for titles such as Genshin Impact, Pokémon Go, Cuphead, Beatsaber, and Call of Duty: Mobile.

What does this mean?

Unity says the move will make it easier for users to build and distribute their games on Windows and Xbox.

In addition, the companies said the partnership will help developers create 3D environments for both gaming and non-gaming scenarios.

In a blog post announcing the news, Microsoft Corporate Vice President Game Creator Experiences and Ecosystem Sarah Bond predicted that real-time simulation will become "central to every industry ranging from e-commerce to energy, manufacturing to medical and more".

"Unity is building a platform-agnostic, cloud-native solution that meets the wide-ranging needs of all developers from enterprise through citizen creators," said Bond. "By giving creators easy access to RT3D simulation tools and the ability to create digital twins of real-world places and objects, Unity is offering creators an easy path to production of RT3D assets, whether for games or non-gaming worlds."

Unity's game engine is already in use in many non-gaming use cases, for example, Boeing has used Unity's software to trial augmented reality-based inspections and testing.

The partnership comes as Microsoft is seeing cloud-based gaming become a larger part of its business model.

In its latest earnings report, the tech giant said Xbox Cloud Gaming has now been used by over 10 million people.

Neither Microsoft nor Unity mentioned when the first collaborative project using the integration is set to land.

• Running out of disc space? Checkout our guide to the best cloud storage 

Microsoft has announced that it wants to extend the lifespan of its cloud servers from four years to six years. 

In a recent earnings call, Amy Hood, Microsoft's chief financial officer, said she believes the move could save the tech giant $3.7 billion in its 2023 financial year alone, and around $1.1 billion in the first quarter.

This is likely unwelcome news for firms such as Intel, AMD, and Ampere, which make millions every year supplying the hardware fueling Microsoft's cloud infrastructure.

Why the move?

Hood attributed the changing speed of the updates to investments in its software "that increased efficiencies in how we operate our server and network equipment as well as advances in technology have resulted in lives extending beyond historical accounting useful lives".

Microsoft isn't the only big tech firm that is currently extending the life cycles of its hardware.

Both Amazon Web Services and Google both added another year to the lifecycle of their cloud infrastructures in February 2022. 

Anything that helps Microsoft deal with the surging demand for cloud hosting may be welcome news. 

Technology news website The Information reported that two dozen Azure data centers internationally are operating with limited capacity, citing two Microsoft managers with knowledge of the issue.

It's not all doom and gloom for data center hardware providers, however, with Meta just one of the big tech firms pouring money into new data centers.

The firm spent $7.7 billion on capital expenditure in its latest quarter, with data center-related investments representing a pretty big chunk of the spending.

• Want to move your data away from public cloud? Checkout our guide to the best bare metal servers

Via The Register