The U.S. government issued an enforcement letter to Anthropic, compelling the company to disable its Fable 5 and Mythos 5 models just before the weekend. The letter, sent by the Commerce Department on June 15, 2026, invoked an obscure export control directive, restricting non-Americans from accessing the models. Anthropic stated it believed the directive was linked to a potential bypass of the models’ guardrails but noted the letter lacked specific details. The company complied by shutting down both models to all customers, effectively halting their use in the U.S. This swift action by the Trump administration demonstrated the government’s ability to enforce compliance without court approval, raising concerns about the potential for similar interventions in the AI industry. The incident highlights the growing tension between U.S. tech firms and government oversight, with implications for the broader cybersecurity and AI sectors.
The situation escalated over the weekend, with Axios reporting that the dispute stemmed from personality differences between Anthropic and the Trump administration rather than a technical issue with the models. New details emerged about a private paper shared by Anthropic with cybersecurity researcher Katie Moussouris, which described an alleged guardrail bypass in Fable 5. Moussouris argued that the bypass should not have triggered an export control measure, as the behavior described in the paper could not be meaningfully fixed without weakening the model’s defenses. She and other experts criticized the directive as hasty and misguided, calling it a dangerous move that undermines U.S. cybersecurity capabilities.
The incident has drawn comparisons to past export control policies, such as those from the 2010s, which inadvertently restricted legitimate cybersecurity research. However, the Trump administration’s directive appears to be retaliatory, with speculation about potential political motivations. Justin Hendrix of Tech Policy Press noted that the move could signal to foreign governments that U.S. AI companies are not trustworthy without government oversight. The White House has not confirmed the reason for the directive, leaving uncertainty about the broader implications for the AI industry. The episode underscores the risk of government overreach in regulating advanced technologies.
Source: techcrunch
