Anthropic Releases Claude Fable 5 to Enterprise and Paid Users

Anthropic has released Claude Fable 5, a new Mythosclass AI model, to enterprise customers and paid subscribers. The launch comes about two months after the company first limited access to Mythos because of concerns about how such advanced systems could be misused. The company said the broader release is possible because of added safeguards designed to block responses in highrisk areas such as cybersecurity and biology. If a user asks a dangerous question, the model can refuse and redirect to a safer Anthropic model instead. Anthropic said the model performs strongly on software engineering and knowledge work benchmarks, and that the launch fits its longterm goal of scaling Mythosclass systems more widely. The release also comes as Anthropic prepares for a possible IPO and faces growing competition in the AI market.