Anthropic Launches Claude Fable 5, Bars Security Features in Public Mythos
Anthropic launched Claude Fable 5, a Mythos-class AI model with enhanced capabilities, to enterprise customers and paid subscribers two months after its initial private rollout. The company simultaneously released a public edition of Mythos with guardrails that disable its cybersecurity and software-flaw detection functions.
1. Claude Fable 5 Rollout
On June 9, Anthropic began offering Claude Fable 5 to its enterprise clients and paid subscribers, marking the first general release of its Mythos-class model outside of an April private preview. This version retains advanced natural language understanding and reasoning capabilities that outpace predecessors.
2. Guarded Public Mythos Edition
Simultaneously, Anthropic launched a public Mythos edition with robust guardrails that disable its cybersecurity and software-flaw detection functions. These constraints aim to mitigate misuse risks in sensitive domains while providing developers with powerful AI tools for safer applications.
3. Commercial and Security Implications
By segmenting access between enterprise and public users, Anthropic positions itself to capture subscription revenue from high-value clients while controlling downstream risks. The tiered model could accelerate commercial adoption and influence competitive dynamics in the AI platform market.
