Tensions Escalate Between White House and Anthropic Over AI Export Controls

The Trump administration and AI developer Anthropic are locked in a deepening conflict following a government directive that forced the company to disable access to its Fable 5 and Mythos 5 models. Defense Secretary Pete Hegseth, who previously labeled Anthropic a supply chain risk and banned its tools from the Pentagon, publicly criticized the firm on social media, reinforcing the administration's hardline stance. The dispute centers on an export control order that Anthropic claims was issued with little warning and lacked clear justification regarding national security threats.

Anthropic contends that the government’s directive was an overreach, noting that the company had already received explicit approval for the deployment of Fable 5 after undergoing pre-release testing. While the administration reportedly cited concerns over potential "jailbreaking" vulnerabilities, Anthropic argues that these minor flaws are common across the industry and do not warrant such drastic intervention. The company warned that the administration’s opaque, rapid-fire enforcement approach could effectively paralyze the development and deployment of frontier AI models across the entire sector.

This confrontation highlights a growing friction between the White House and leading AI labs regarding the balance of power in model oversight. While Anthropic’s CEO, Dario Amodei, has publicly supported voluntary government testing and third-party auditing, the company insists that regulatory actions must be transparent, fair, and grounded in technical evidence. As Anthropic prepares for a potential public offering, this regulatory volatility poses significant questions about the future of AI governance and the extent to which the government can intervene in the operations of private technology firms.