US Lifts Export Block on Claude Mythos 5: 100 Institutions Cleared for Access

The US government just flipped the switch on Claude Mythos 5's export ban, allowing over 100 institutions, including major companies and government agencies, to access this AI model. This comes after a two-week standoff with the Trump administration, which had put the brakes on Anthropic's AI due to security concerns. But while Mythos 5 gets the green light, its consumer twin, Fable 5, remains in limbo.

Commerce Secretary Howard Lutnick broke the news in a letter to Anthropic's compute chief, Tom Brown. The decision means no license is needed for Mythos 5 exports to entities listed in a specific annex. This reversal comes after senior Anthropic staff made their way to Washington to negotiate terms. The initial block was sparked by Amazon's alarm over Fable 5, fearing it could be misused if jailbroken.

Until now, Mythos 5 was part of Project Glasswing, a vulnerability-hunting initiative active across 150 organizations in more than 15 countries. The model impressed by spotting flaws in classified systems during government tests. Now, with the export block lifted, it's ready to roll out to trusted partners. Meanwhile, Fable 5, once the most powerful AI tool accessible via subscription, stays offline. The AI regulatory world is evolving, with new gatekeeping systems taking shape thanks to recent executive orders.

For the crypto world, this means institutions might be better equipped to handle AI-driven threats and opportunities, but retail users are left out in the cold with Fable 5's indefinite delay. The clear winners here are the big players who now have a advanced tool at their disposal. Consumers, meanwhile, will need to wait and see if Fable 5 ever joins the party.