Anthropic Unveils Fable 5 AI Model to Public

A new version of the AI tool known as Claude Mythos 5 has been released to the public after the company said it was too powerful for open access. With safeguards and user limits in place, Anthropic claims the release is safe for regulated use.

When Mythos first entered preview in April, it raised serious worries about security risks, prompting concerns from finance and government leaders. The model’s ability to exploit or hack computer systems made it a marked concern for some experts.

Security teams that used Mythos have already reported more than 10,000 critical flaws in their systems, illustrating the model’s sheer power. The company now says the 150 preview groups will gain full access to “Claude Mythos 5,” which does not have the same constraints on bio‑ or cyber‑security.

Anthony’s COO Jack Clark said that the rapid ascent of AI tools suggested the need for a “brake pedal” – a way for the public to slow development if needed. He notes that the industry currently feels like it is only pressing the gas.

Anthropic is also slated to become a public company soon, with a private valuation approaching $1 trillion. The broader access rollout is expected to extend beyond a small group of cyberdefenders and infrastructure providers in the near future.

For more details, see the original BBC article at Finance ministers raise concerns over Mythos AI.