Anthropic releases Claude Fable 5, Mythos AI with safety

On Tuesday, June 9, 2026, Anthropic released Claude Fable 5, the first publicly available version of its most advanced "Mythos" class of artificial intelligence models. The launch represents a significant step in making frontier AI capabilities broadly accessible, but comes with novel safety restrictions designed to prevent misuse in high-risk domains such as cybersecurity and biology.

The company stated that Fable 5 outperforms all its previous publicly released models across software engineering, knowledge work, vision, and scientific research, with its advantage growing on longer and more complex tasks. Early third-party testing cited by Anthropic showed the model achieving top scores on analytical benchmarks and excelling in areas like UI design and game coding. A notable example provided was from Stripe, which reported the model completed a codebase-wide migration across 50 million lines of Ruby code in one day, a task estimated to take a team over two months.

However, due to the model's enhanced capabilities, Anthropic has implemented a safeguard system. When Fable 5 detects user requests related to cybersecurity, biology, chemistry, or model distillation, it automatically...