BBC reported on June 10, 2026 that Anthropic released Claude Fable 5—a publicly available version of Claude Mythos, an AI program the company previewed privately in April and initially framed as potentially too dangerous for general release because of its ability to exploit or hack computer systems.

What BBC says Anthropic launched

  • Claude Fable 5 is described as a public edition of Mythos with safeguards and user limitations in place.
  • Anthropic said on Tuesday that "releasing a model this capable comes with risks" and that "Fable's capabilities exceed those of any model we've ever made generally available."
  • The company also said roughly 150 groups that previewed Mythos will now access Claude Mythos 5, which carries fewer limitations on cybersecurity and biology depending on an organization's approved uses.
  • Preview users have reported finding more than 10,000 critical security flaws in their systems, according to Anthropic statements cited by BBC.
  • Access to Mythos 5 remains limited to a "small group of cyberdefenders and infrastructure providers", with Anthropic saying it intends to expand access through a broader trusted access program.
  • BBC notes Fable and Mythos are essentially the same underlying model with different safeguards and access tiers, and can work unattended on human commands for longer periods than prior Claude models.

Policy, government, and market context in the piece

  • Canadian Finance Minister François-Philippe Champagne told the BBC in April that attention on Mythos was warranted in part because "it's the unknown, unknown."
  • U.S. government agencies have been testing Mythos despite Anthropic's ongoing lawsuit against the U.S. Department of Defense over government use of its AI tools, BBC reported.
  • Anthropic is expected to become a public company soon, with private valuation nearing $1tn (£747bn), the article said.
  • Co-founder Jack Clark told BBC Newsnight last week that AI capability is expanding so rapidly the industry needs a public brake: "Right now, it's like the AI industry has a gas pedal, but it doesn't have a brake pedal."
  • BBC also records skepticism from some observers about how much of the capability hype is marketing spin, while others worry the tool could pose financial security risks.

Primary source: BBC — Version of AI tool 'too powerful for public' released to public (June 10, 2026).