More than 100 cybersecurity experts and industry leaders have urged the Trump administration to withdraw restrictions on Anthropic’s latest artificial intelligence models, warning that the move could ultimately benefit US adversaries rather than strengthen national security.

In a letter sent Sunday, executives and experts, including representatives from Adobe and Nvidia, called on the government to lift export-control directives affecting Anthropic’s advanced AI systems, Fable 5 and Mythos 5, and adopt a more transparent, science-based approach to assessing AI-related risks.

OpenAI faces multistate probe over user safety concerns ahead of IPO

Anthropic said Friday it had taken the models offline to comply with the directive, despite arguing that the government’s concerns over potential security risks did not justify the restrictions. The company previously limited access to Mythos 5 because of its ability to identify and exploit software vulnerabilities at a level exceeding human cybersecurity experts.

The signatories acknowledged the models’ strong cybersecurity capabilities but noted that similar functions are available through other leading and open-source AI systems. They warned that restricting access to advanced defensive tools without clear justification could weaken US cyber defenses while rivals, particularly China, continue to narrow the technological gap.

The dispute comes amid growing tensions between Anthropic and the Trump administration. The company has challenged Pentagon efforts to label it a supply-chain risk following disagreements over military applications of its AI technology and concerns about autonomous weapons and domestic surveillance.