Anthropic is racing to restore access to its most advanced artificial intelligence (AI) models after a White House export-control order restricted their use on Friday, with senior company officials now in Washington seeking a resolution, according to an Axios report.
Why The Government Stepped In
Anthropic received the directive from U.S. authorities, requiring the company to block all foreign nationals from accessing Fable 5 and Mythos 5, including foreign-national employees working at the company.
According to an Axios report, citing a source close to the company, administration officials believe Anthropic has "not engaged in a serious manner", even though the company has been holding virtual meetings with White House officials since initial outreach began Friday.
Anthropic did not immediately respond to Benzinga's request for comment.
Anthropic Pushes Back
On Friday, the company argued that the reported jailbreak exposed only a limited number of previously known vulnerabilities and did not represent a broad failure of the models' safety protections.
Anthropic said no researcher has demonstrated a universal jailbreak capable of bypassing the safeguards built into Fable 5 and Mythos 5.
The company described the directive as a "misunderstanding" and said it is working to restore access while complying with the government's order.
Larger Fight Over AI
The timing is particularly significant because Fable 5 and Mythos 5 were introduced only days ago and were promoted as Anthropic's most advanced AI systems after extensive testing involving government agencies, researchers and internal teams.
The battle over Fable 5 and Mythos 5 is now shaping up as an early test of how much influence governments will wield over the next generation of frontier AI models.