US-based artificial intelligence company Anthropic is currently in discussions with the Trump administration in an effort to reverse newly imposed restrictions on access to its most advanced AI systems, according to a report published by The Wall Street Journal on Monday, News.Az informs.

The restrictions, which were introduced on Friday, prohibit foreign governments, foreign companies, and individual users from accessing Anthropic’s most advanced models, named Fable 5 and Mythos 5. The measures were put in place citing national security concerns.

In response to the restrictions, Anthropic officials reportedly traveled to Washington over the weekend in an attempt to reach a resolution with US authorities.

According to the report, the discussions included US Commerce Secretary Howard Lutnick, National Cyber Director Sean Cairncross, Anthropic co-founder Tom Brown, and Sarah Heck, who serves as the company’s head of public policy.

Anthropic stated that the government had instructed it to suspend access for all foreign nationals, which includes employees of the company working within the United States.

In its statement, the company said it believes the US directive was linked to concerns that its Fable 5 model could potentially be “jailbroken,” meaning its safeguards might be bypassed in ways that could enable harmful or unintended uses. Anthropic, however, defended the effectiveness of those safety safeguards.

The company first revealed the existence of its Mythos model in April, but it did not publicly release it at the time due to concerns about its capability to identify vulnerabilities in software systems.

More recently, last Tuesday, Anthropic launched Mythos 5 for a limited group of users, alongside Fable 5, which was made available as a public version but with additional safety restrictions and safeguards in place.

News.Az 

By Nijat Babayev