Anthropic launches Claude Fable 5 following Wall Street disruption
The release of Claude Fable 5 marks the first public availability of the company’s advanced AI family, two months after a private iteration caused significant market turbulence.
Anthropic has officially released Claude Fable 5, the first public iteration of its Mythos-class artificial intelligence model family, according to a report by CNBC. The launch arrives two months after a private rollout of the more powerful Claude Mythos 5 caused significant disruption on Wall Street, prompting the company to implement stricter controls before making the technology widely available.
The company stated that the broad public release is enabled by new safeguards designed to block or defer responses in specific high-risk areas. These measures address previous concerns that the Mythos class possessed capabilities in cybersecurity and biology that were deemed too dangerous for unrestricted public distribution. The release represents a shift in Anthropic’s strategy, balancing rapid market availability with stringent security protocols.
Claude Fable 5 is characterised by exceptional performance in software engineering, knowledge work, and vision tasks. Testing indicates that the model’s lead over competitors grows as tasks become longer and more complex. In internal testing, 95 per cent of Claude Fable 5 sessions ran entirely on Fable responses without falling back to Claude Opus 4.8, demonstrating a high degree of stability in standard operations.
To manage risk, the model utilises a tiered safety protocol. It defaults to Claude Opus 4 for general operations but defers to Claude Opus 4.8 in high-risk domains. This architecture allows the system to handle complex queries while maintaining control over sensitive areas, such as those previously identified as concerns regarding biological and cyber capabilities.
While the public version is now available, the original private model, Claude Mythos 5, remains accessible to specific institutional partners with lifted safety restrictions. This dual-track approach allows Anthropic to serve high-value institutional clients with fewer limitations while offering a safer, albeit still powerful, version to the general market.
