Microsoft has canceled most internal Claude Code licenses in its Experiences and Devices division, effective June 30, 2026. The decision comes just six months after the company introduced the tool to engineers in December 2025. The reason is financial: token-based billing consumed the annual AI budget far ahead of schedule, and the company is redirecting its engineers to GitHub Copilot CLI, Microsoft’s own AI coding assistant.
The Experiences and Devices division is responsible for Windows, Microsoft 365, Outlook, Teams, and Surface hardware, making it one of the highest-volume engineering organizations in the company. When Claude Code was introduced, engineers adopted it rapidly, apparently preferring Anthropic’s tool over Microsoft’s own offering. That adoption rate, combined with token-based pricing that charges per request, produced costs that escalated far faster than budget projections anticipated.
Uber’s situation provides a parallel. The company reportedly burned through its entire 2026 AI budget of $3.4 billion in just four months after deploying Claude Code to 5,000 engineers. Monthly usage rates reached 84 to 95% by April 2026, and per-engineer API costs ranged between $500 and $2,000 monthly. Uber is now internally warning that it may exhaust its remaining AI budget before the year ends.
The Microsoft decision is notable for several reasons. First, it is unusual for a company to cancel a tool primarily because engineers used it too much rather than too little. Claude Code delivered enough demonstrable value that engineers chose it over the in-house alternative at scale. Second, it exposes a structural problem with consumption-based pricing for AI coding tools: the more productive the tool, the faster the budget disappears. Third, it creates pressure on Anthropic to develop enterprise pricing structures that give companies cost predictability, something that flat-fee or seat-based models provide and token-based billing does not.
The timing of the cancellation, aligned with Microsoft’s June 30 fiscal year end, is deliberate. Ending the licenses now prevents token costs from rolling into the new fiscal year and gives the company a defined window to build governance frameworks before any future AI coding tool rollout.
For sales teams tired of cold leads, slow customer responses, and manual processes, Dapta is the ultimate tool.
Dapta is the leading platform for creating AI sales agents specifically designed to increase inbound lead conversion. Respond to your leads in less than a minute with voice AI and WhatsApp that converts.
If you want your team to sell more while AI handles the complex stuff, you have to try it.
For enterprise technology buyers in Latin America and globally, the Microsoft and Uber situations are a concrete warning: before deploying AI coding or agent tools at scale, establish usage governance frameworks. That means per-team budgets, usage monitoring, and a clear policy on acceptable consumption. The companies that deploy AI tools without those guardrails are likely to face the same outcome: genuine productivity gains, unsustainable costs, and a difficult decision about whether to cut a tool that teams actually value.
