Anthropic in talks with Trump administration to reverse AI model restrictions: Report
Cybersecurity experts call for lifting curbs on Fable 5, Mythos 5, warning restrictions could weaken US AI leadership, cyber defense capacity
US artificial intelligence firm Anthropic is holding talks with the Trump administration to reverse restrictions imposed on its most advanced AI models, The Wall Street Journal reported Monday.
The restrictions, introduced Friday, bar foreign governments, companies and individuals from accessing Anthropic's Fable 5 and Mythos 5 models on national security grounds.
Anthropic officials traveled to Washington over the weekend to seek a resolution, according to the report.
The discussions involved Commerce Secretary Howard Lutnick, National Cyber Director Sean Cairncross, Anthropic co-founder Tom Brown and Sarah Heck, the company's head of public policy.
Anthropic said the government ordered it to suspend access for all foreign nationals, including employees working at the company inside the US.
In a statement, the company said it believes the US directive was related to concerns that Fable 5 could potentially be "jailbroken" to bypass safeguards designed to prevent harmful uses. Anthropic defended the effectiveness of those safeguards.
Anthropic first disclosed Mythos in April but did not release it publicly because of concerns over its ability to identify software vulnerabilities.
Last Tuesday, the company launched Mythos 5 for selected users alongside Fable 5, a public version with additional safeguards.
Experts urge reversal of restrictions
The move prompted criticism from cybersecurity executives, researchers and technology experts, who sent an open letter to the Trump administration calling for the restrictions to be lifted.
The signatories argued that Fable 5 and Mythos 5 do not possess unique capabilities unavailable in other advanced AI systems and warned that limiting access could weaken both US leadership in artificial intelligence and cyber defense efforts.
They also called for future AI risk assessments to be conducted through a more transparent and scientific process.
Previous tensions over military use
The restrictions come amid broader tensions between Anthropic and the US government over military applications of advanced AI systems.
Anthropic has resisted allowing its technology to be used for certain military purposes, citing concerns about surveillance of US citizens and autonomous weapons.
The disagreement escalated earlier this year after the Pentagon pressed the company to permit broader use of its AI models by the US military. President Donald Trump subsequently directed federal agencies to halt the use of Anthropic technology, while the Pentagon designated the company a procurement-security risk.
Anthropic later filed lawsuits against the US government, alleging officials had unlawfully retaliated against the company over its position on AI safety.
The latest restrictions on Fable 5 and Mythos 5 have further intensified debate in Washington over how to regulate frontier AI models without undermining US technological competitiveness.
