Microsoft CEO Nadella faults Anthropic's Claude Fable limit

Microsoft CEO Satya Nadella has publicly criticized the safety and access restrictions placed on Anthropic's advanced AI model, Claude Fable, calling its frequent refusal of user requests illogical for a creation tool. Nadella made the remarks during an internal meeting with engineers working on Microsoft's Copilot AI software on Wednesday, July 15, 2026, according to a transcript of his comments.

Nadella focused on the model's pattern of rejecting or redirecting user prompts, questioning the utility of an AI assistant designed for creative work that behaves as if it is under heavy editorial control. He argued that such restrictive safeguards do not make practical sense for users. His criticism extends beyond a single feature to a broader economic concern about concentration in the AI industry. Nadella stated that it cannot be sustainable for only a handful of companies to control the essential computing resources, or "token capital," required for advanced AI, forcing everyone else to effectively rent access.

Anthropic introduced Claude Fable 5 in early June 2026 as one of its most capable models, incorporating stricter safety measures than previous versions. According to the...