Anthropic PBC has stopped access to its most advanced artificial intelligence models, including Mythos, after an unprecedented order by the Trump administration to keep the technology out of the hands of all foreign nationals.
The US government directed Anthropic to suspend access to the Fable 5 and Mythos 5 models by any foreign national “whether inside or outside the United States,” due to national security concerns, the company said in a statement.
A US official confirmed the Commerce Department sent the letter, Bloomberg News reported. The compnay since then has shut off access to both systems to all customers.
This is the first time when the US government taken such sweeping measures to rein in foreign access to frontier AI models developed by an American company. Both the Trump and Biden administrations have limited access abroad to other consequential technologies such as semiconductors and supercomputers, and some have argued the merits of blocking access to AI models. However, curbs on the software itself have raised constitutional and commercial concerns.
According to a report by Axios, Commerce Secretary Howard Lutnick on Friday sent a letter to Anthropic CEO Dario Amodei mentioning that export controls will be levied on Mythos 5 and Fable 5 models in any location outside of the U.S. and to all foreign persons within the country.
The letter was sent after the administration tried to get Anthropic to pause releasing the latest models but was unsuccessful, it added, The letter mentioned that a license will be required for the export, re-export or domestic transfer of those Anthropic models. Additionally, the company will have to submit an additional application for individually validated licenses. In case Anthropic fails to comply with the driective, financial and civil penalties are likely to be levied, the report said.
ALSO READ: Explained: What Makes Fable 5, Mythos 5 Prized Assets Of United States
According to Anthropic, the US government issued the order after finding out that it's possible to “jailbreak,” or bypass the guardrails, of Fable 5, a recently released version of Mythos that the company blocked from carrying out cybersecurity tasks.
“We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people,” Anthropic said in its website post. “If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers.”
Amazon.com researchers had conducted jailbreak research that showed some vulnerabilities in Anthropic's model, according to a report in the Wall Street Journal.
Both US government and Amazon were in contact about the vulnerability before the curbs were imposed, people familiar with matter told Bloomberg. Amazon CEO Andy Jassy was also engaged in those exchanges, one of the people said. He even raised concerns to senior US officials earlier, Information reported.
The government's move to so widely restrict access to a set of AI models threatens to set a narrative for all major AI model developers including OpenAI, Alphabet Inc.'s Google and Meta Platforms Inc. Major industry players such as Nvidia Corp. head Jensen Huang and OpenAI CEO Sam Altman have always encouraged the US government to instead promote worldwide adoption of American AI systems.
“For anyone who was naive and perhaps hoping that this leverage wouldn't be exerted, it's a massive wake-up call,” Aidan Gomez, the co-founder of Cohere Inc., a Nvidia Corp.-backed AI startup, said Saturday in an interview. “No one can deny it any more.”
The latest directive runs counter to earlier statements, as well as an executive order recently signed by US President Donald Trump, which indicated the government would not pursue a licensing regime for model reviews.
It further threatens to escalate long-standing tensions between Anthropic and some within the Trump administration. The AI developer clashed with the Pentagon over the use of its technology for military and surveillance purposes earlier this year. The government called the company a US supply-chain risk as a result of the issue and instructed US agencies to phase out the use of its products.
Anthropic first released its Mythos model in April to a very limited group of companies and institutions, cautioning that its ability to find cybersecurity vulnerabilities made it too risky to distribute more widely.
There were indications that the limited release was working to ease tensions between Anthropic and the Trump administration.The US government was preparing to make a version of Mythos available to major federal agencies in April, according to reports.
Mythos also boosted the Trump administration's efforts on AI policy, which included the recent executive order that urged for voluntary model review. It explicitly said that nothing in it should be construed as creating a mandatory licensing regime.
David Sacks, Trump's former AI czar and current co-chair of the President's Council of Advisers on Science and Technology, said that Anthropic refused to fix a jailbreak of the guardails in its Fable model.
“The Admin's hope now is that Anthropic remediates the safety issue, the export control is lifted, and Fable goes back into general release,” he wrote in a post on X. “The Admin wants all of this to happen as soon as possible. It is frankly bewildered that Anthropic hasn't wanted to comply with safety requests that it previously said were its highest priority.”
The latest government restriction is clashing with a race among US AI developers to deliver the most advanced models and prove to their investors that the technology can turn a profit. Both OpenAI and Anthropic are seeking initial public offerings as soon as this year, following SpaceX's own historic IPO.
Essential Business Intelligence, Continuous LIVE TV, Sharp Market Insights, Practical Personal Finance Advice and Latest Stories — On NDTV Profit.