ANTHROPIC RELEASES PUBLIC VERSION OF MYTHOS AI MODEL WITH RESTRICTIONS

Anthropic has made a version of its Mythos artificial intelligence model available to the general public. The release, named Fable 5, marks the first Mythos-class model to receive wide public access. Anthropic promoted Fable 5 as a tool for writing and debugging software code, answering complex research questions, and analysing images. The Mythos lineup was unveiled in April but had been restricted to a small group of partner institutions for months over cybersecurity concerns.

Alongside the public release, Anthropic continues to offer an unrestricted version, Claude Mythos 5, to companies and organisations already enrolled in its Project Glasswing programme. That group expanded in early June to approximately 200 organisations across more than 15 countries. Anthropic has limited broader access to the Mythos family on the grounds that the technology can rapidly identify vulnerabilities in critical infrastructure, including banking platforms and power grids. When Project Glasswing launched, some critics accused Anthropic of overstating the threat to generate attention, though companies that tested Mythos subsequently endorsed its capabilities.

The US government, which had previously been in a legal dispute with Anthropic, tested the Mythos model over security concerns. The White House has since established an arrangement to assess the most powerful AI models from leading companies before they are publicly released. Anthropic confirmed the Project Glasswing programme is expected to expand beyond its current 200 organisations. The release of Fable 5 represents the first time Mythos-class technology has been accessible to the general public.