Microsoft Bans AI For Employees As It Costs More Than Humans, Big Tech Faces AI Cost Reality Check

US based tech giant Microsoft, which is also considered one of the biggest backers of artificial intelligence in the world, has quietly started pulling back on AI tools for its own engineers. The reason is not that technology stopped working. The bill just got too big to ignore. This comes at a time when Nvidia’s own VP has openly admitted that running AI costs his team more than paying its human employees. The two stories together are forcing a hard question that nobody in the tech world really wants to answer right now: is AI actually worth what it costs?

What Microsoft Actually Did 

Microsoft is canceling thousands of internal Claude Code licenses across its Experiences and Devices group, the team behind Windows, Microsoft 365, Outlook, Teams, and Surface, with a 30th June 2026 cutoff. Engineers are being moved to GitHub Copilot CLI, Microsoft’s own tool, even though many of them preferred Claude Code. 

To be clear, this is not a full ban on AI. Anthropic’s models are not being banned at Microsoft. Claude will still be available inside Copilot CLI. Only the Claude Code interface is going away internally. But the message it sends is hard to miss. Microsoft pumped billions into AI, told its engineers to go use it, and then watched the costs shoot up faster than anyone planned for.

Microsoft began canceling most of those direct Claude Code licenses and redirecting engineers toward GitHub Copilot CLI instead. This was not about souring on Anthropic as a partner. What changed was internal math. When a tool becomes popular at scale under a token-based pricing model, the costs do not grow in a straight line. They can explode.

The Nvidia Admission That Stopped Everyone 

Around the same time, a senior Nvidia executive made a remark that spread fast. Bryan Catanzaro, vice president of applied deep learning at Nvidia, told Axios that for his team, the cost of compute is far beyond the costs of the employees, making AI more expensive than human labour.

That is a striking thing to hear from someone at Nvidia, the company that builds the very chips powering all this AI. If even Nvidia’s internal teams are finding AI more expensive than people, it tells you something real about where the economics currently stand.

Uber Ran Out of AI Budget in Four Months 

Microsoft and Nvidia are not alone in this. Uber’s CTO Praveen Neppalli Naga told The Information in April that the firm had already burnt through its entire 2026 AI coding tools budget in just four months. Four months. The whole year’s budget, gone before summer.

What the Numbers Say 

A 2024 MIT study found AI automation is economically viable in only about 23% of jobs, with humans still cheaper in the remaining 77%. Despite unclear productivity gains and high costs, big tech companies have committed around $740 billion to AI-related expenses this year, a 69% jump from 2025.

This increase in spending has also come alongside more layoffs in the tech sector. According to data from Layoffs.fyi, there have been more than 92,000 layoffs in tech in 2026 so far across nearly 100 companies. Companies are replacing people with AI that costs more than the people it replaced. That is the situation right now.

Is This a Problem or Just Growing Pains? 

Not everyone sees this as a crisis. Keith Lee, an AI and finance professor at the Swiss Institute of Artificial Intelligence, told Fortune that what we are seeing is a short-term mismatch, and that the cost of using AI has remained less efficient than human labor owing to hardware and energy raising operating costs for providers. He expects the economics to shift as infrastructure improves, and model costs come down.

Also Read: CPUs Becoming Important Again As AI Inferencing Demand Rises: AMD Expects Strong Growth In AI-Focused Processors