Anthropic is meeting with U.S. officials to resolve a dispute over export restrictions on its latest AI models, Fable 5 and Mythos 5. The Trump administration's directive, citing national security concerns about potential bypasses, led Anthropic to disable access globally.
Government Directive and Immediate Action
Senior technical staff from Anthropic are scheduled to meet with U.S. Department of Commerce officials in Washington, D.C., on Monday to address a dispute over export restrictions on the company's latest AI models, Fable 5 and Mythos 5. The Trump administration ordered Anthropic to suspend access to these models for foreign nationals, citing national security concerns. In response, Anthropic disabled access globally.
The Trump administration issued an export control directive on Friday, requiring Anthropic to block foreign nationals from using Fable 5 and Mythos 5. The company complied by disabling access to the models worldwide. The government cited concerns over a potential method to bypass safeguards, allowing the models to identify software vulnerabilities.
Background of the Dispute and Company Response
Anthropic has been in discussions with U.S. officials since Friday, with daily virtual meetings. The company argues that the security bypass did not constitute a full "jailbreak" of the models' guardrails. However, the administration remains unconvinced, pointing to reports that Amazon discovered a method to bypass Mythos 5's security features.
Anthropic released a statement on Friday, announcing the suspension of access to the models following the government's directive. The company had previously worked with government agencies to test the models before their release. The dispute marks the latest tension in Anthropic's relationship with the U.S. government, which has labeled the company a national security risk.
Government's Stance and Legal Context
White House AI czar David Sacks criticized Anthropic CEO Dario Amodei for prioritizing model availability over safety. The administration claims Anthropic was less cooperative in resolving the security concerns. Defense Secretary Pete Hegseth supported the restrictions, citing ongoing litigation between Anthropic and the government over a previous supply chain risk designation.
The restrictions come amid broader debates over AI regulation and national security. The Trump administration's actions reflect growing concerns about foreign access to advanced AI technologies. Anthropic's lawsuit against the government over the supply chain risk designation remains ongoing, adding legal complexity to the dispute.