On Oct 5, 2026 Meta and Microsoft announced they are cutting internal use of Claude, Anthropic’s flagship chatbot. At Microsoft the annual spend on Claude, once projected at over $1 billion, was slashed by more than a third. Executives told staff to favor GitHub Copilot and OpenAI’s models instead. The cloud division’s per‑employee budget fell from $100,000 to about $10,000 a month.
Meta’s Claude Code users dropped from roughly 60,000 to 30,000. In a recent 28‑day period the company spent over $105 million on Claude Code alone, even as it pushes its own tools such as Muse Code and MetaCode. The cuts follow earlier AI cost‑saving steps announced in June.
Why the pull‑back matters
For employees, the change means they will be using Microsoft’s Copilot or OpenAI’s models rather than Anthropic’s Claude for coding and office‑type tasks. For Anthropic, losing two of its biggest enterprise customers could tighten its revenue stream and accelerate the shift toward competing AI products from Meta and Microsoft.
Why it matters
Employees at the two firms will now rely on internal tools or OpenAI’s services, changing which AI they interact with daily. Anthropic may see a dip in revenue as its biggest enterprise customers turn away, signaling a shift in the competitive landscape.
