Microsoft’s CEO criticized Anthropic for imposing restrictions on the use of its Fable model, arguing that enterprises should independently develop low-cost AI models.

According to CNBC, Microsoft CEO Satya Nadella criticized Anthropic’s usage restrictions on its Fable model during an internal meeting for Copilot engineers, describing the company’s approach of rejecting certain requests as “excessively restrictive editing control” and calling it unreasonable. Anthropic’s support page notes that when users ask for content related to large language model (LLM) development, the platform may switch to an older model version to respond. The firm previously stated that new security measures rolled out after Fable 5’s relaunch could lead to more false positives, incorrectly flagging harmless requests for blocking. Microsoft earlier announced a $5 billion investment in Anthropic, with the latter committing to spend $30 billion on Azure. Some of Microsoft’s Copilot products also use Anthropic’s models, so Nadella’s comments involve its key partner and customer. Nadella also emphasized that enterprises should be able to develop custom models at lower costs and control their internal data, rather than allowing a small number of companies to monopolize AI token resources. He said: “We can’t have just two global companies holding token capital while everyone else is forced to rent. That doesn’t make economic sense.”