Meta and Microsoft reportedly curb employee use of Claude

Meta and Microsoft are reportedly trying to reduce employees’ use of Anthropic’s Claude as they steer workers toward their own AI tools. The reasons for the shift are not clear, and the spending figures cited have not been independently confirmed.

Microsoft had asked employees to limit Claude use and spend more time with Copilot and other Microsoft tools, according to The Information. Its spending on Claude was estimated to be on course to reach at least $1 billion this year, the report said.

Meta has reportedly halved the number of employees using Claude Code, from 60,000 earlier this year to 30,000. Its internal alternative, MetaCode, has passed 30,000 users. The company was also set to spend billions on internal AI use before imposing token limits, spending controls and other monitoring, according to the source.

The shift could affect Anthropic’s business. In an IPO prospectus, the company warned that it “may experience material revenue losses or business disruptions attributable to these events,” Reuters reported. Reuters also reported that Anthropic was slated to go public at a valuation of more than $2 trillion. The source says the US Department of Defense had designated Anthropic a national security supply chain risk.