Cybersecurity Chiefs Demand Trump Admin Loosen Anthropic AI Curbs

A significant directive from the Trump administration preventing foreign nationals from using Anthropic's latest artificial intelligence models, Fable 5 and Mythos 5, has drawn strong criticism from a group of over 100 cybersecurity executives and experts. This coalition argues that such restrictions could inadvertently aid U.S. adversaries rather than hinder them. In compliance with the government directive, Anthropic announced Friday that it has taken its advanced AI models offline, though the San Francisco-based company stated its belief that the government's concerns regarding potential security issues did not warrant these extreme measures.

Anthropic had previously limited access to its most cutting-edge technology, including the more advanced Mythos model, to select customers due to its exceptional ability to identify and exploit computer vulnerabilities, surpassing even human cybersecurity experts. The company had engaged in prior discussions with the White House regarding the capabilities of these models. The letter, signed by leaders from prominent companies like Adobe and Nvidia, urged the U.S. government to rescind the export control directives and to establish a transparent and scientific process for future AI risk assessments.

The experts' letter highlighted that while Anthropic’s Mythos models are indeed effective at discovering software flaws and weaponizing exploits, they are not "uniquely good at these tasks." Many signatories indicated they regularly utilize other foundation and open-source models for security audits and training. They warned that removing superior cyber defense capabilities without justifiable cause is perilous, especially as America's adversaries rapidly advance their own AI technologies. The letter specifically mentioned that China's AI models are "only months behind the best American models," and suggested that the Chinese government likely possesses private capabilities beyond public knowledge.

This directive represents the U.S. government's most substantial action to date in restricting access to advanced AI models. Anthropic had released Fable, a more limited version of Mythos, widely just last week. The Commerce Department has not yet responded to requests for comment regarding the directive. This move by the administration followed President Donald Trump's executive order, signed ten days prior, which sought to create a framework for federal government vetting of national security risks associated with advanced AI systems before their public release, with developer participation being voluntary.

Tensions between the safety-conscious Anthropic and the Trump administration have been apparent. Anthropic has consistently advocated for robust guardrails in AI development to mitigate potential risks while maximizing economic and national security benefits for the U.S. A previous contract dispute with the Pentagon saw Defense Secretary Pete Hegseth attempt to declare Anthropic a supply chain risk. This unprecedented action against a U.S. company, which Anthropic has challenged in federal courts, stemmed from the company's desire for assurances that its technology would not be used in fully autonomous weapons or for the surveillance of Americans. Hegseth, conversely, maintained that the company must permit any uses the Pentagon deemed lawful.