Anthropic suspends AI model access after U.S. security order

Traders Union
All News
Financial News
Anthropic AI model suspension

Highlights

Anthropic suspends access to its Mythos 5 and Fable 5 AI models globally after a U.S. government order citing a foreign national jailbreak risk.
President Donald Trump's executive order establishes a voluntary framework for agency review of frontier AI systems but does not grant formal model release blocking powers.
Anthropic's legal dispute with the U.S. government over Pentagon supply-chain risk designation intensifies policy tensions, raising sector-wide concerns about future AI deployment standards.

Government order halts model rollout

As reported by Financial Times, commerce secretary Howard Lutnick directs Anthropic on Friday afternoon to limit access to foreign nationals after the government discovers a jailbreak in the model's safeguards. Anthropic says the practical effect of that order is a full shutdown of access, forcing it to switch off the services everywhere rather than only outside the U.S.

The company says it has withdrawn Mythos 5 and Fable 5 from all users, even though the Trump administration has access to Anthropic's frontier Mythos model for several weeks before the release. Federal agencies are experimenting with the model ahead of its launch, while some officials raise concerns that it could be used to exploit cybersecurity vulnerabilities in critical infrastructure, including the global financial system.

President Donald Trump recently signs an executive order that stops short of giving the government formal powers to block a model release, while setting up a voluntary framework for agencies to inspect frontier systems before wider deployment. Anthropic says the latest action rests on a misunderstanding and argues that a narrow jailbreak finding should not trigger a recall of a commercial model used by hundreds of millions of people.

Legal dispute deepens policy tensions

Anthropic says the government should be able to block unsafe deployments only through a statutory process that is transparent, fair, clear and grounded in technical facts. It argues that if the same standard were applied across the sector, new model launches by frontier AI providers would effectively come to a halt.

The clash adds to an already strained relationship between Anthropic and the Trump administration as both sides continue to disagree over how and when powerful AI systems should be deployed. Anthropic is suing the U.S. government over its Pentagon designation as a supply-chain risk after a dispute related to military use of its models, and the company has also faced criticism from members of Trump's coalition, including former AI tsar David Sacks.

Trump has separately suggested that AI companies should donate equity to the government, though that proposal has not moved forward. The White House does not immediately respond to a request for comment.

In our earlier article on the procurement crackdown over “Made in America” claims on GSA Advantage!, we explained how the SBA and GSA removed foreign-made product listings after findings that China-based sellers had marketed them as American-made. We also noted the administration’s broader push to steer federal purchasing toward domestic suppliers, alongside new SBA support measures for U.S. manufacturers.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.