Key Takeaways:
The US government ordered Anthropic to suspend access to its two most powerful AI models, citing national security risks from a reported jailbreak.
The Commerce Department on Friday ordered Anthropic to block foreign access to its Mythos 5 and Fable 5 models, escalating Washington's campaign to treat frontier AI systems as national security assets.
"The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees," the company said in a statement. Anthropic received the order at 5:21 p.m. ET on June 12 and disabled access immediately.
Commerce Secretary Howard Lutnick sent a letter to Anthropic CEO Dario Amodei saying the models would require a license for export, re-export or domestic transfer to any foreign person. An administration official said the decision followed a claim by another company that it could jailbreak Mythos, alarming officials about potential national security risks. Anthropic reviewed the demonstration and called the vulnerability minor, noting that other publicly available models, including OpenAI's GPT-5.5, can discover similar flaws without any bypass.
The directive places Anthropic in a bind: the company is on a Pentagon blacklist deeming its models too dangerous for government use, and now under a Commerce licensing regime deeming them too dangerous for foreign access. The restriction could affect revenue from international clients and sets a precedent that may slow frontier AI deployments across the industry.
Anthropic emphasized that Fable 5 launched with stronger safeguards than any previously deployed model. Before release, the company worked with the US government, UK AISI and multiple third-party teams to red-team the safeguards for thousands of hours. No tester found a universal jailbreak capable of bypassing Fable 5's protections across a wide range of cyber capabilities, the company said. Anthropic adopted a defense-in-depth approach combining narrow safeguards, monitoring and 30-day data retention for Mythos-class models.
The company argued that pulling a commercial model deployed to hundreds of millions of users over a narrow vulnerability sets a problematic precedent. "If applied across the industry, this standard would essentially halt all frontier AI model deployments," Anthropic said.
The Trump administration earlier this month released an executive order to test the most advanced AI models before deployment. However, the order is voluntary and explicitly avoids a licensing regime — something White House chief AI adviser David Sacks secured to avoid what he considers regulatory capture of the biggest labs. An administration official said Trump "does not want to hurt the industry and wants innovation to continue."
Anthropic has a partnership with the Center for AI Standards and Innovation at Commerce for pre-deployment testing. The company said it plans to share more technical details over the next 24 hours and is working to restore access to Fable 5 and Mythos 5 as soon as possible.
For investors, the export controls introduce regulatory risk into Anthropic's valuation as it prepares for a public debut. The restriction limits the addressable market for its most advanced models and could push international clients toward non-US AI competitors. Anthropic's ability to resolve the dispute with the Commerce Department in the coming weeks will determine whether the disruption remains a short-term setback or becomes a structural drag on its growth.
This article is for informational purposes only and does not constitute investment advice.
