Kategorie

Microsoft experienced a significant service disruption across its Microsoft 365 services on Monday, affecting core applications including Microsoft Teams and Exchange Online. The outage left users globally unable to access collaboration and communication tools critical to consumers as well as enterprise workflows.

In a series of updates posted on X through the official account of Microsoft 365 Status, Microsoft acknowledged the incident and confirmed that it was actively investigating user reports of service impact. The incident was tracked under the identifier MO1096211 in the Microsoft 365 Admin Center.

Minutes after initial acknowledgement, Microsoft initiated mitigation steps and reported that all services were in the process of recovering. “We’ve confirmed that all services are recovering following our mitigation actions. We’re continuing to monitor recovery,” the company said in an update.

Roughly an hour later, Microsoft posted another update, saying, “Our telemetry indicates that all of our services have recovered and that the impact is resolved.”

“The Microsoft outage that disrupted Teams, Exchange Online, and related services was ultimately caused by an overly aggressive traffic management update that unintentionally rerouted and choked legitimate service traffic. According to Microsoft’s official post-incident report, the faulty code was rolled back swiftly, but not before triggering global access failures, authentication timeouts, and mass user logouts,” said Sanchit Vir Gogia, chief analyst and CEO at Greyhound Research.

Microsoft did not immediately respond to a request for comment.

Not an isolated incident

This incident adds to a growing number of high-profile cloud service disruptions across the industry, raising questions about the resilience of hyperscale infrastructure and the impact on cloud-dependent enterprises. In the last 30 days, IBM Cloud services were disrupted thrice, and a Google Cloud outage just last week impacted over 50 services globally for over seven hours.

Microsoft, in particular, has experienced a steady stream of service disruptions in recent months, exposing persistent fault lines in its cloud infrastructure.

In March this year, the outage disrupted Outlook, Teams, Excel, and more, impacting over 37,000 users. In May, Outlook suffered another outage, which was attributed to a change that caused the problem.

According to Gogia, this sustained pattern reveals architectural brittleness in Microsoft’s control-plane infrastructure — especially in identity, traffic orchestration, and rollback governance — and reinforces the urgent need for structural mitigation.

Costly outages call for contingency planning

Given the complexity and global scale of hyperscale cloud infrastructures, outages remain an ongoing risk for leading SaaS platforms, including Microsoft 365. More so for enterprises that operate in hybrid and remote work environments, threatening business continuity.

Such outages can lead to loss of productivity and disrupted communications, depending on the applications they affect as well as the extent of the outage. This could mean a loss of thousands of dollars to potentially millions of dollars for some, explained Neil Shah, vice president of research, Counterpoint.

Manish Rawat, analyst, TechInsights, said industry estimates suggest that IT downtime can cost mid- to large-sized enterprises between $100,000 and $500,000 per hour, depending on their sector and the criticality of operations. “For large organizations, even a brief 2–3 hour outage could result in millions in lost productivity, reputational harm, and serious operational setbacks, especially in high-stakes sectors like finance, healthcare, and manufacturing,” he said.

Given the recent incidents involving Microsoft 365 services alone, experts believe that enterprises must reduce their overdependence on Microsoft 365. “Organizations should adopt robust contingency plans that include alternative communication tools, offline access to critical documents, and a comprehensive incident response framework,” said Prabhu Ram, VP for industry research group at CMR.

Grammarly is reinventing itself as a platform of generative AI (genAI) agents that go beyond grammar recommendations.

The company is building always-on AI technologies that follow users across work applications and can coordinate projects, write documents and automate workflows.  The growth is partially fueled by a cash infusion of $1 billion from General Catalyst last month.

Users will be able to have deeper conversations with, and get recommendations from, a variety of AI tools that draw context from documents and action. The company hopes the tools will attract power users who want more than automatic recommendations.

“We’re going to be able to give you much more feedback than, ‘Here are correct words,’” Noam Lovinsky, chief product officer at Grammarly, told Computerworld. “We’ll be able to give you feedback from experts that you care about. We’ll be able to help you right from start to finish.”

For example, an agent could access Zoom transcripts from candidate interviews and create draft scorecards.

“Maybe at the end of that process, Grammarly says, ‘Actually, I have this agent. And if you want, I can like create a draft for you of every single score card the minute you get off of the Zoom,’” Lovinsky said.

Grammarly can take that further in its Coda team workspace tool. It can automate that post-Zoom workflow by generating a table of all interviews, linking transcripts, and generating draft responses. Users can review and refine the drafts before sending the data to hiring systems such as Greenhouse.

“That is almost like the common interface layer by which all the agents are going to get to show up. They’ll come to your applications the same way that Grammarly does today,” Lovinsky said.

Lovinsky offered another example of how Grammarly tools will work with popular collaboration tools from Slack and Atlassian. “If you’re writing a status update for a project, we’ll actually know the latest things that have been said in Slack about that project and the latest Jira tickets that have been filed, so we can help you coalesce those things and create a good update.”

Grammarly claims 40 million users and supports 500,000-plus applications and websites. Its tools typically slip into the user interface without disturbing the flow of work, which is what the company wants to continue.

“You just install it and it just works, and you just go do your thing and we show up in the right ways and in the right moments,” Lovinsky said.

The genAI technologies will include a companion that can go deeper into context, which could appeal to power users. “If you want more than just what shows up in the underlines, if you want deeper work with more back and forth, we’ll have a companion that opens up and works with every application,” Lovinsky said.

Grammarly appears to have thought about what technology it has that it could build on to offer a differentiated genAI tool, said Nancy Gohring, senior research director for AI at IDC. “What it landed on was the platform it had already created that allowed the original Grammarly application to work across third-party applications,” she said.  

Grammarly’s advantage comes from leveraging its existing platform as a way to offer a range of agents that work across the third-party apps already deployed in the enterprise, Gohring said.

But it will also compete with numerous companies delivering similar tools. Microsoft and Google already provide document drafting and automation AI agents. Users rely on large-language models (LLMs) for grammar and error correction.

But with LLMs, users must cut, paste, and prompt to get answers. Grammarly wants to ease that burden by working within the user interface and automatically understanding the right context.

“What I want to do is create an interface that doesn’t require you to prompt and re-prompt until you get your output,” Lovinsky said. “It just works.”

Grammarly builds its own LLMs, but also uses commercial AI providers. Users cannot explicitly choose their model, but that limitation could change.

“I think we’re going to change that because there are more sophisticated users,” Lovinsky said. “What they really like is that we bring it inline.”

The company plans to focus on tools for knowledge workers in the wider market and not target specific domains. “You’re not going to see us do a lot in coding, data analysis, or design work,” Lovinsky said.

There are already many rivals on the market that aim to allow for agent development, management and collaboration, IDC’s Gohring said.

“Grammarly will need to clearly articulate how it’s different and how it fits adjacent to the others,” she said.

The relationship between OpenAI and Microsoft is under growing strain amid extended talks over OpenAI’s restructuring, with OpenAI reportedly considering antitrust action over Microsoft’s influence in the partnership.

OpenAI leaders have considered alleging that Microsoft engaged in anticompetitive practices during their collaboration, a move that could prompt a federal investigation, WSJ reported.

The ChatGPT maker is reportedly exploring the option of urging regulators to examine its contractual relationship with Microsoft, along with a public campaign.

Meanwhile, The Information reported that OpenAI is seeking to give Microsoft a roughly 33% stake in its reorganized for-profit unit in exchange for relinquishing rights to future profits.

OpenAI also wants to revise existing contract clauses that grant Microsoft exclusive cloud hosting rights and to exclude its planned $3 billion acquisition of AI startup Windsurf from terms that give Microsoft access to OpenAI’s intellectual property, the report added.

These developments threaten to disrupt one of the most closely watched alliances in the AI sector.

A potential antitrust complaint by OpenAI could heighten regulatory scrutiny of major AI-cloud partnerships and lead enterprise customers to reevaluate risks tied to vendor lock-in and control over core infrastructure.

Microsoft, a major investor since 2019, supports OpenAI through Azure and powers tools like Microsoft 365 Copilot with its models.

However, tensions between OpenAI and Microsoft have been simmering in recent months, with occasional public clashes.

OpenAI has also been trying to reduce its dependence on Microsoft by turning to Google Cloud for additional computing power, while Microsoft has been working to lessen its own reliance on OpenAI by integrating alternative AI models into its Copilot platform, according to Reuters.

Impact on enterprises

A potential regulatory review may weaken enterprise confidence in adopting or expanding the use of Copilot and related tools, particularly in heavily regulated sectors such as healthcare and financial services.

“Over the short to long term, enterprises could face service disruptions, compatibility issues, or increased costs as vendors adjust their business models in response to changes in the partnership or service offerings,” said Prabhu Ram, VP of the industry research group at CyberMedia Research.

OpenAI models currently power Microsoft Copilot. But with growing innovation from rivals like DeepSeek, both firms appear to be preparing for a more independent path.

“The rate at which AI is advancing, especially given what DeepSeek has demonstrated, suggests that being locked into a single model is no longer a prudent strategy for Microsoft,” said Neil Shah, VP of research and partner at Counterpoint Research. “Enterprises will need to prepare for AI tools and platforms that are diverse in capability, modular, and scalable.”

For OpenAI, partnerships with Oracle Cloud and potentially Google Cloud will help scale its models further in enterprise deployments, particularly in the public sector, where Google is working to expand its presence.

“In the end, most cloud and AI providers will need to support multiple models and adopt modular integration to give enterprises more choice,” Shah said. “This way, they avoid becoming a one-trick pony and can select models based on their strengths, future development roadmaps, and alignment with specific use cases.”

Threat intelligence researchers are warning of hackers breaching multiple U.S. companies in the insurance industry using all the tactics observed with Scattered Spider activity. [...]

Generative AI (genAI) poses a classic IT dilemma. When it works well, it is amazingly versatile and useful, fueling dreams that it can do almost anything. 

The problem is that when it does not do well, it might deliver wrong answers, override its instructions, and pretty much reinforce the plotlines of every sci-fi horror movie ever made. That is why I was horrified when OpenAI late last month announced changes to make it much easier to give its genAI models full access to any software using Model Context Protocol (MCP).

“We’re adding support for remote MCP servers⁠ in the Responses API, building on the release of MCP support in the Agents SDK⁠,” the company said. “MCP is an open protocol that standardizes how applications provide context to LLMs. By supporting MCP servers in the Responses API, developers will be able to connect our models to tools hosted on any MCP server with just a few lines of code.”

There are a large number of companies that have publicly said they will use MCP, including those with  popular apps such as PayPal, Stripe, Shopify, Square, Slack, QuickBooks, Salesforce and GoogleDrive.

The ability for a genAI large language model (LLM) to coordinate data and actions with all of those apps — and many more —certainly sounds attractive. But it’s dangerous because it allows access to mountains of highly sensitive compliance-relevant data — and a mistaken move could deeply hurt customers. MCP would also allow genAI tools to control those apps, exponentially increasing risks.

If the technology today cannot yet do its job properly and consistently, what level of hallucinogens are needed to justify expanding its power to other apps?

Christofer Hoff, the CTO and CSO at LastPass, took to LinkedIn to appeal to common sense. (OK, if one wanted to appeal to common sense, LinkedIn is probably not the best place to start, but that’s a different story.) 

“I love the enthusiasm,” Hoff wrote. “I think the opportunity for end-to-end workflow automation with a standardized interface is fantastic vs mucking about hardcoding your own. That said, the security Jiminy Cricket occupying my frontal precortex is screaming in terror. The bad guys are absolutely going to love this. Who needs malware when you have MCP? Like TCP/IP, MCP will likely go down as another accidental success. At a recent talk, Anthropic noted that they were very surprised at the uptake. And just like TCP/IP, it suffers from critical deficiencies that will have stuff band-aided atop for years to come.”

Rex Booth, the CISO at identity vendor SailPoint, said the concerns are justified. “If you are connecting your agents to a bunch of highly sensitive data sources, you need to have strong safeguards in place,” he said. 

But as Anthropic itself has noted, genAI models do not always obey their own guardrails. 

QueryPal CEO Dev Nag sees inevitable data usage problems. 

“You have to specify what files [the model] is allowed to look at and what files it is not allowed to look at and you have to be able to specify that,” Nag said. “And we already know that LLMs don’t do that perfectly. LLMs hallucinate, make incorrect textual assumptions.”

Nag argued that the risk is — or at least should be — already well known to IT decision makers. “It’s the same as the API risk,” Nag said. “If you open up your API to an outside vendor with their own code, it could do anything. MCP is just APIs on steroids. I don’t think you’d want AI to be looking at your core financials and be able to change your accounting.”

The best defense is to not trust the guardrails on either side of the communication, but to give the exclusion instructions to both sides. In an example with the model trying to access Google Docs, Nag said, dual instructions are the only viable approach.

“It should be enforced at both sides, with the Google Doc layer being told that it can’t accept any calls from the LLM,” Nag said. “On the LLM side, it should be told ‘OK, my intentions are to show my work documents, but not my financial documents.’”

Bottom line: the concept of MCP interactiveness is a great one. The likely near-term reality? Not so much.

As companies move from testing out generative AI tools and models into real-world use — also known as inference— they’re having trouble predicting what that use will lead to in terms of cloud costs, according to a new report from analyst firm Canalys .

“Unlike training, which is a one-time investment, inference represents a recurring operational cost, making it a crucial constraint on the path to commercializing AI,” said Canalys senior director Rachel Brindley in a statement. “As AI moves from research to large-scale deployment, companies are increasingly focusing on cost-effectiveness in inference, comparing models, cloud platforms, and hardware architectures such as GPUs versus custom accelerators.”

According to Canalys researcher Yi Zhang, many AI services rely on usage-based pricing models that charge per token or API call; that makes it difficult to predict costs when scaling up usage.

“When inference costs are volatile or excessively high, companies are forced to limit usage, reduce model complexity, or restrict implementation to high-value scenarios. As a result, the broader potential of AI remains underutilized,” said Zhang.

A high-severity vulnerability in ASUS Armoury Crate software could allow threat actors to escalate their privileges to SYSTEM level on Windows machines. [...]

Email accounts of several Washington Post journalists were compromised in a cyberattack believed to have been carried out by a foreign government. [...]

Kali Linux 2025.2, the second release of the year, is now available for download with 13 new tools and an expanded car hacking toolkit. [...]

Zoomcar Holdings (Zoomcar) has disclosed via an 8-K form filing with the U.S. Securities and Exchange Commission (SEC) a data breach incident impacting 8.4 million users. [...]

Apple stealthily introduced Apple Sparse Image Format (ASIF), a new sparse disk image format for Apple Silicon, at WWDC; among other features, it might also help Macs remain the best PCs on which to run Windows.

That somewhat counter-intuitive claim is because the new format dramatically improves the efficiency with which Apple Silicon Macs run virtual machines (VMs) by boosting read/write performance. ASIF also makes significant improvements to how Macs handle storage for VMs, and is likely to support third-party virtualization tools once it ships.

What is Apple Sparse Image Format (ASIF)?

Set to appear with macOS Tahoe later this year, ASIF lets files transfer more efficiently between hosts or disks, “because their intrinsic structure doesn’t depend on the host file system’s capabilities,” according to Apple’s developer website. “The size the ASIF file takes on the file system is proportional to the actual data stored in the disk image.”

[ Related: Apple WWDC 2025: News and analysis ]

ASIF replaces the currently used format, which occupies the same amount of disk space as the allocated portion of the disk. What this means is that when you allocate 10GB of disk space to a VM you immediately sacrifice 10GB of space, no matter how much data the virtual machine contains; with ASIF, the volume will only occupy as much space as it contains. In other words, you can allocate a large quantity of disk space to enable optimal VM performance but sacrifice only as much actual space as the VM contains. 

“These space-efficient images can be created with the diskutil image command-line tool or the Disk Utility application and are suitable for various uses, including as a backing store for virtual machines storage via the Virtualization framework,” Apple explained.

The company says users should migrate their VM storage images from the existing RAW format to ASIF to benefit from the improved file transfer performance between the host Mac and the disk. 

Faster and highly performant

Eclectic Light was first out the gate with news about Apple’s new tech, publishing a first look alongside test results to show how much faster it is in use than standard sparse imaging technologies it is.

The test results show that ASIF gives Disk images on Apple Silicon devices near-native SSD speeds. That matters whenever you are moving data around, and is particularly important when running Linux or Windows in virtual machines. It means you should experience significant performance benefits, further reinforcing the Mac as the best platform for Windows.

Eclectic Light noted that in some cases an encrypted sparse image (UDSP) stored on the fast SSD of a current Mac might only write files at up to an unimpressive 100 MBps. That report comes with receipts, sharing extensive test data to show that even encrypted ASIF files read and write data far faster than the sparse file formats Macs use today. That’s will mean much more moving forward as on-device encryption becomes even more essential to personal data protection as government mandated back doors are identified and abused.

It’s not just about virtual machines. For general storage, it seems highly probable the new format will also improve the performance of FileVault. That’s because right now encrypted sparse images are used to secure a user’s home directory in FileVault, so better performance and storage management have implications there. (I also speculate that the new format might have implications in how Apple efficiently provides and encrypts future LLM services via Private Cloud Compute.)

Additional WWDC coverage:

• WWDC: For developers, Apple’s tools get a lot better for AI
• Apple’s AI Revolution: Insights from WWDC
• WWDC 2025: What’s new for Apple and the enterprise
• WWDC: What we know so far about Apple’s Liquid Glass UI
• WWDC first look: How Apple is improving its ecosystem.
• Apple WWDC 2025: News and analysis

You can follow me on social media! Join me on BlueSky,  LinkedIn, and Mastodon.

Microsoft has shared a workaround for a known issue that causes the classic Outlook email client to crash when opening or starting a new message. [...]

In a bid to deliver secure, compliant generative AI (genAI) to business teams without the need for a cloud service, Uptime Industries last week unveiled Lemony. It’s a turnkey stackable device that comes preloaded with multiple large language models (LLMs) and can serve up to five users per node, connected directly to a PC or to a LAN with no internet connectivity. As business needs expand, multiple nodes can be connected into a cluster, with automatic failover. Lemony says a four-node cluster can support up to 50 users and comes with six pre-loaded genAI models.

IBM is working with the company to deploy its Granite AI models on Lemony nodes. Other available LLMs include Llama-3.1, Llama-3.2, and Mistral. In addition, JetBrains is integrating its coding models and tools into the Lemony node to allow software developers to leverage its intelligent development features. (For maximum performance, each node contains a neural processing unit (NPU), an AI accelerator cluster and a CPU.)

Other pre-loaded functions include retrieval-augmented generation (RAG), and the ability to create AI assistants to help with tasks such as analyzing documents. 

Data is loaded onto a Lemony node in one of three ways, with only a knowledge graph retained, said Uptime CEO and cofounder Sascha Buehrle. “We are uploading, analyzing, indexing, and deleting the data,” he said. 

Users can also connect to their data via an API, with which the data is indexed, or connectors that integrate directly with business applications.

One early customer, Alexander Göbel, legal tech officer at Niederer Kraft Frey AG in Zurich, Switzerland, called setup quick and easy. “You can be up and running with an on-premises solution within minutes rather than within days/weeks,” Göbel said via email. Transferring documents to the device, where they’re indexed for use via RAG, is done by uploading via the Lemony browser.

 “We are currently working on a SharePoint connector to make the process of uploading even simpler, as the latest version of the node provides for a limited API,” Göbel said.

Because the nodes are not connected to the internet, Lemony updates are provided quarterly via individually keyed encrypted USB keys. Each key will only work with its designated node. The update also resets a secure timer to ensure that the user’s subscription is still valid; if not, the node locks, with data fully encrypted.

Lemony offers a two-week free trial, and Uptime says it already has more than 300 customers in Switzerland, Germany, the UK, and the US. Subscriptions start at $499 per month for a single node accessed by up to five users, billed annually; the setup includes the node, software, apps for Windows and Mac, and technical support. 

Cautious optimism

Analysts found the concept appealing, and were cautiously optimistic about the device itself.

“Uptime is tapping into a real need for any regulated industry,” said Matt Kimball, vice president and  principal analyst for Datacenter Compute & Storage at Moor Insights & Strategy. “And in Europe this is most companies. What we effectively have is a genAI appliance. 

“If I am a CIO or an IT professional at a smaller law firm,” Kimball said, “I immediately see the value in this Lemony platform…. I can see Lemony being attractive at the departmental level [at a larger firm] or for an SMB that values/requires data privacy. And the ability to use AI without the need for IT is super interesting.” 

“The on-prem AI edge is an underserved segment,” said Gartner Vice President Analyst Chirag Dekate. “Most genAI infra[structure] today assumes cloud-first. There is an opportunity for localized solutions, especially if latency, cost, or compliance are concerns. If Uptime provides automated [machine learning] ops, energy optimization, and support for open-source models, it might reduce the complexity barrier enough to attract mid-sized enterprises and public sector clients. Global expansion of AI regulations will make ‘keep your AI local’ more attractive in the next two to three years.”

Wyatt Mayham, lead AI consultant at Northwest AI Consulting, agreed. “We work with clients who refuse to put sensitive data in the cloud, even if it’s Azure + OpenAI, which never touches the public web or trains the models,” he said. “Clients often think they want true on-prem, but actually building an on-prem setup with GPUs, model hosting, orchestration, and RAG infrastructure is expensive, high maintenance, and usually way overkill for what they actually need. 

“This actually looks like a solid middle ground,” Mayham said. “It’s not full-scale enterprise infra, but it gives small teams a path to locally run LLMs, stay compliant, and avoid the cloud.”

“We don’t consider Lemony.ai as a replacement for all cloud-based AI systems,” Göbel said. “For ‘commodity data’ with lower confidentiality requirements, a turnkey cloud solution remains to make sense to us (in this case, access to internet information may be required, too). However, we are dealing with a lot of very sensitive and confidential information for which cloud solutions are not options. As a result, depending on the specific use cases, Lemony.ai and cloud-based solutions work very well in tandem.”

Lemony’s success will be influenced by how it’s positioned in the market, Dekate said. ”Uptime’s Lemony AI strategy will be limited if it’s positioned as a general-purpose AI appliance. But if Uptime focuses on narrow verticals with repeatable workloads — like retail, energy, or industrial monitoring — it may gain traction,” he said.

But there are “fundamental challenges in a packaged AI in a box experience (independent of a vendor): Packaging doesn’t solve talent gaps. Just because it’s in a box doesn’t mean it’s plug-and-play for everyone. AI model management, updates, and troubleshooting are still hard. Without model agility, customers may view it as a closed system, limiting experimentation and extensibility. [And] if Uptime isn’t controlling its hardware supply chain or relies on commodity boards, this may be hard to differentiate long-term.”

Kimball sounded a cautionary note: “I will counter with this one thing: if I were still in IT, I am not sure I would be allowing ‘AI appliances’ to populate my network.  If you thought shadow IT was bad with the cloud — holy moly!”

Law enforcement authorities from six countries took down the Archetyp Market, an infamous darknet drug marketplace that has been operating since May 2020. [...]

Despite sizeable tech layoffs over the past two years, a tech talent gap persists — especially for those trained on implementing and using generative artificial intelligence (genAI) tools. Consultancy McKinsey & Co. now projects that demand for AI-skilled workers will outpace supply by two-to-four times, a skills gap likely to continue at least until 2027.

That echoes what consultancy Deloitte wrote in a recent report. It found that corporate leaders continue to rate critical talent shortages as one of their greatest fears, even as job-seekers report despair about their hiring prospects. “And yet neither side seems prepared to address it,” Deloitte said in its report.

A ManpowerGroup survey of 40,413 employers in 42 countries found that 74% of employers still struggle to find skilled talent, with only 16% of execs confident in their tech teams and 60% citing the skill gaps as a key barrier to digital strategies. Along the same lines, Bain & Co. found that 44% of corporate leaders say limited in-house expertise has slowed AI adoption, with demand for AI skills rising 21% annually since 2019 and a shortage of talent lasting another two years.

The good news? Pay for AI skills continues to increase, growing 11% a year since 2019, according to Bain & Co. Workers with AI skills such as prompt engineering command a 56% wage premium (up from 25% last year), suggesting the value they bring, according to PricewaterhouseCoopers (PwC).

PwC said its data “does not show job or wage destruction from AI,” it shows growth across AI-exposed roles — even highly automatable ones. AI is boosting expertise, allowing workers to take on higher-level tasks. according to Joe Atkinson, PwC’s Global chief AI officer.

PricewaterhouseCoopers

“AI is at the forefront of corporate transformation, but without the right talent, businesses will struggle to move from ambition to implementation,” Sarah Elk, head of AI research for Bain & Co.’s Americas group, said in a statement. “Executives see the growing AI talent gap as a major roadblock to innovation, limiting businesses’ ability to scale and compete in an AI-driven world.”

While the gap has always been a challenge, gaining experience is harder than ever with AI taking over various work tasks, remote work weakening apprenticeships, and rising job complexity that require broader skills, Deloitte said.

The AI skills gap is driven by the rapid growth of AI technologies and the increasing demand for adoption across industries, according to Kelly Stratman, Ernst & Young’s global ecosystem relationships enablement leader. “Currently, 50% of enterprises with more than 5,000 employees have adopted AI solutions, and even more are considering doing so. At the same time, job postings requesting AI skills increased by 2000% in 2024 alone.”

By 2030, companies are expected to spend $42 billion a year on genAI projects such as chatbots, agents, research, writing, and summarization tools.

Key AI skills in short supply include prompt engineering, programming, and bias handling. Just as vital are soft skills such as adaptability, critical thinking, and emotional intelligence to ensure responsible, ethical AI use, according to Stratman.

PwC’s new AI Jobs Barometer shows demand for AI skills growing, even as the US job market slows. In response, PwC recently launched two AI tools that offer career development through tailored training and an AI coach that adapts to each employee’s goals and projects.

Bain & Co. projects AI job demand could reach up to more than 1.3 million in the US over the next two years, while the number of skilled workers available is on track to hit less than 645,000 — implying the need to reskill up to 700,000 US workers. “Companies navigating this increasingly competitive hiring landscape need to take action now, upskilling existing teams, expanding hiring strategies, and rethinking ways to attract and retain AI talent,” Bain’s Elk wrote.

The first response to the skills gap is to take an honest assessment of your organization and align AI projects with core business goals, “because really, this isn’t just a question of AI readiness, it’s about digital, data and AI readiness,” tech consultancy Thoughtworks wrote in a new report.

The key takeaways from Thoughtworks’ report:

• Strategic alignment matters: 61% of leaders have a mature tech strategy compared to 19% of late adopters, showing its impact on digital and AI success.
• Continuous improvement is essential: 93% see room for tech ecosystem improvement; 77% of leaders seek major changes.
• Tech leadership boosts ROI: 53% of leaders report positive ROI, outpacing all other groups.

Justin Vianello, CEO of US technology talent training firm SkillStorm, said a shortage of qualified talent — especially in cloud, cybersecurity, and AI — is a bigger barrier to hiring than AI automation replacing jobs. Organizations struggle to find candidates with the right skills, certifications, and clearances, Vianello said, referring to shortages for government agencies.

While AI can boost productivity by handling routine tasks, it can’t replace the strategic roles filled by skilled professionals, Vianello said. To avoid those kinds of issues, agencies — just like companies — need to invest in adaptable, mission-ready teams with continuously updated skills in cloud, cyber, and AI.

The technology, he said, should augment – not replace — human teams, automating repetitive tasks while enhancing strategic work.

Success in high-demand tech careers starts with in-demand certifications, real-world experience, and soft skills. Ultimately, high-performing teams are built through agile, continuous training that evolves with the tech, Vianello said.

“We train teams to use AI platforms like Copilot, Claude and ChatGPT to accelerate productivity,” Vianello said. “But we don’t stop at tools; we build ‘human-in-the-loop’ systems where AI augments decision-making and humans maintain oversight. That’s how you scale trust, performance, and ethics in parallel.”

High-performing teams aren’t born with AI expertise; they’re built through continuous, role-specific, forward-looking education, he said, adding that preparing a workforce for AI is not about “chasing” the next hottest skill. “It’s about building a training engine that adapts as fast as technology evolves,” he said.

Microsoft acknowledged a new issue caused by the June 2025 security updates, causing the DHCP service to freeze on some Windows Server systems. [...]

Navigating the AI talent shortage

The IT layoffs we’ve seen in 2024 and early 2025 are set to continue as companies look to drive efficiencies with AI while bracing for a recession. This wee,k CIO.com reported on company boards pushing CEOs to replace IT workers with AI. It’s compelling reading.  

That’s the bad news for IT professionals. But our readers wanted to understand where the opportunity lies in all this disruption. Many of them asked Smart Answers where the AI skills gap is — and where are the roles going to be. 

The good news is that more than three-quarters of employers say they’re struggling to find the right tech talent, with roles in AI and Machine Learning most prominently cited as gaps. 

Find out: Where is the AI talent shortage most prominent now?  

Why do CISOs quit?

Staying with the challenge of hiring and retaining IT talent, this week, CSO reported that more than half of department heads reporting to CISOs are looking to quit. It’s an irony that only a softening economy is keeping many of them in their jobs.  

But what about the CISOs themselves? Many readers asked Smart Answers why CISOs tend not to stay in their roles for long. Leaning into decades of human reporting, our AI answer service surmises that the issues include high stress levels, personal liability and organizational distance from decision-makers. (Those would make us change jobs, too.)  

Find out: Why is average CISO tenure so short? 

…And why you should stop CISOs quitting

Meanwhile, here’s the problem with all of those CISOs quitting: a good CISO is worth his or her weight in gold — or even more valuable than that. 

This week, CSO published an article in which David Gee argued that good CISOs have highly specialized knowledge that takes significant time and investment to develop for a secure future. So you need to keep them onboard. They are hard to find.  

Why is that? Smart Answers thinks the issue is escalating cyberattacks and the growing sophistication of security threats, including AI-driven attacks, that is driving up demand. More people, with more skills, are required.  

Find out: Why are cybersecurity professionals difficult to hire currently?  

About Smart Answers 

Smart Answers is an AI-based chatbot tool designed to help you discover content, answer questions, and go deep on the topics that matter to you. Each week we send you the three most popular questions asked by our readers, and the answers Smart Answers provides.  

Developed in partnership with Miso.ai, Smart Answers draws only on editorial content from our network of trusted media brands—CIO, Computerworld, CSO, InfoWorld, and Network World—and was trained on questions that a savvy enterprise IT audience would ask. The result is a fast, efficient way for you to get more value from our content. 

ChatGPT's Codex, which is an AI agent that lets you code and delegate programming tasks, is now testing a new feature that lets you choose the best solution. [...]

On June 13, OpenAI began rolling out a new ChatGPT Search update to improve quality as the AI startup challenges Google's dominance. [...]