A version of an artificial intelligence tool that its own maker said was too powerful for public release has been made available to everyone. Claude Fable 5, a stripped-back variant of Anthropic's Claude Mythos — an AI program that rattled technology, finance and government leaders when it was first previewed in April — launched on Tuesday.
Anthropic acknowledged the gamble. “Releasing a model this capable comes with risks,” the company said, adding that Fable’s “capabilities exceed those of any model we’ve ever made generally available.” The version comes with safeguards and user limitations, but the company stopped short of guaranteeing safety.
“Anthropic releases Claude Fable 5 to the public, despite calling its predecessor too dangerous.”
When Mythos was quietly handed to a small group of organisations in April, Anthropic privately warned that the tool was so intelligent it could exploit or hack computer systems. That warning did little to contain the fallout. Canadian finance minister François-Philippe Champagne told the BBC at the time that the attention on Mythos was warranted “because it’s the unknown, unknown.”
Now those same organisations — roughly 150 of them — will get access to the full Claude Mythos 5, which lifts restrictions on cybersecurity and biological uses depending on the group’s purpose, Anthropic said on Tuesday. To date, those early testers have already reported finding more than 10,000 critical security flaws in their systems.
The public release of Fable comes as Anthropic’s private valuation nears $1tn (£747bn), with expectations of an imminent public listing. The company is simultaneously fighting a lawsuit against the US Department of Defense over its refusal to accept government use of its AI tools, yet US agencies have been testing Mythos anyway.
Both Fable and Mythos are essentially the same model, differentiated only by safeguards and access levels. Both can work “unattended” on human commands for longer than any previous Claude models, Anthropic said.
Anthropic co-founder Jack Clark told BBC Newsnight last week that AI capabilities were expanding so fast the company believed the public should have a way to slow the technology’s advance. “You want the option to be able to take your foot off the gas and put your foot on the brake,” he said. Whether Fable is the accelerator or the brake remains an open question.