Inside the fight over Claude Mythos 5

Anthropic received a directive from the US government to suspend access to its AI models, Mythos 5 and Fable 5, for any foreign national. This directive also extended to foreign national employees of Anthropic. The company determined that the only way to comply with this directive was to disable both the Mythos 5 and Fable 5 AI models.

Anthropic took steps to address the situation, including traveling to Washington, DC, in an attempt to influence President Donald Trump’s decision regarding the export control directive. The actions taken by the US government could significantly alter the future trajectory of the entire AI industry.

Mythos 5 and Fable 5 are developed on the same foundational technology as Anthropic’s Mythos Preview model. Mythos 5 was made available to a limited group of government agencies and companies. Fable 5, which included additional safeguards, was initially considered safe for general use. However, a report indicated that the guardrails on Fable 5 may have failed.

The Trump administration contacted Anthropic on Friday around 1 PM ET, giving the company a 90-minute deadline to shut down access to Mythos 5 and Fable 5. The administration stated that if Anthropic did not comply, it would impose export controls under the authority of the US Commerce Department. Anthropic executives began discussions with the White House within 15 minutes of the initial call. CEO Dario Amodei joined these discussions approximately one hour and 15 minutes after the first contact. During these discussions, Amodei communicated with key government officials, including US Treasury Secretary Scott Bessent, Commerce Secretary Howard Lutnick, and National Cyber Director Sean Cairncross.

Anthropic stated that it believed the government was aware of a method to bypass or “jailbreak” Fable 5. The company described this jailbreak as a “potential narrow, non-universal” issue that had been shared with the government by an unnamed entity. Anthropic also noted that this behavior was not exclusive to Fable 5 and that similar capabilities are widely available from other AI models, including OpenAI’s GPT-5.5.

A report suggested that the US government’s concerns stemmed from the possibility that a group linked to China had accessed the technology. These concerns reportedly dated back several weeks and involved a large global telecommunications company that had initially been cleared for access to Mythos Preview. In relation to the issue, David Sacks mentioned an unnamed entity that had alerted the government to the problem, describing it as a “highly credible trusted partner of both Anthropic and the USG who was testing Fable.” Some reports point to Amazon CEO Andy Jassy as the individual who raised concerns with the US government after Amazon researchers conducted red-teaming exercises on Fable 5. Anthropic had access to Amazon’s research paper within days of the export control directive and had been in discussions with Amazon researchers about it.

Anthropic mobilized employees, including Dave Orr, Logan Graham, and Nicholas Carlini, to travel to Washington, DC. Reports indicated that Anthropic had repeatedly made communication errors with the administration. The company had been relying on Mythos to help recover from several high-profile disagreements with the US Department of Defense. Anthropic had previously briefed the administration on Fable 5, and the US Department of Commerce had conducted pre-deployment testing without expressing any concerns at that time. Anthropic had maintained close collaboration with government agencies since the release of Mythos Preview.

The Trump administration’s stance on AI safety has become more uncertain. Cybersecurity leaders have cautioned that restricting access to Mythos 5 and Fable 5 could provide China with a significant advantage in the AI race. The President’s decision has prompted international calls for alternatives to American AI systems and has effectively halted the deployment of a major US AI company’s new flagship model. On Sunday, a public letter signed by technology and cybersecurity executives urged the repeal of restrictions on Fable 5. The letter stated that if regulations are to be implemented, they should be based on “scientific evaluations developed with input from industry and academia.”

Posted in Uncategorized
Warning: Undefined array key 0 in /var/www/html/wp-content/themes/highlights-news/single.php on line 32 Warning: Attempt to read property "term_id" on null in /var/www/html/wp-content/themes/highlights-news/single.php on line 32

Leave a Reply

Your email address will not be published. Required fields are marked *