A “Build with Claude” poster at Anthropic’s Code with Claude developer conference in San Francisco on May 6, 2026.Don Feria/AP
On Friday night, the AI giant Anthropic said that the US government had ordered it to suspend foreign nationals, including employees, from all use of its most advanced products.
To comply with the Friday directive, the company announced that it disabled access to Fable 5 and Mythos 5, the latest models of Claude, for all customers.
Anthropic stated that the government cited national security concerns but did not provide further details. The company says its newest technology has enhanced software engineering and visual understanding compared to previous iterations. But Anthropic has also acknowledged potential concerns, releasing a preview model in April to just a few industry partners to test for capabilities to use it to create hacking tools. Claude Fable 5 is the first publicly available version of the Mythos model, and the company said it has established “guardrails” such as blocking answers to questions on cybersecurity, biology, and chemistry.
The Trump administration barred all federal agencies from using Anthropic products in February. That same day, Trump called Anthropic “a radical left, woke company” amid his feud over it being unwilling to permit the military to use its technology. At the time, CEO Dario Amodei said that the US government’s demands—namely, mass surveillance of Americans and fully autonomous weapons—would allow it to violate the company’s safeguard policies.
As my colleagues Anna Merlan and Abby Vesoulis pointed out in March, the US military previously used Anthropic’s Claude for “intelligence assessments, target identification and simulating battle scenarios” to prepare for its initial strikes on Iran.
Anthropic has positioned itself as the ethical AI company, a significant contributor to its rapid ascent to the top of the industry especially as the public has increasingly disapproved of AI development. The company filed for an initial public offering earlier this month, and SpaceX’s success so far since it entered the stock market on Friday—which made founder Elon Musk a trillionaire—could be an encouraging sign for it and its major competitor OpenAI.
Meanwhile, other countries, like China and the United Arab Emirates, are pushing for “sovereign AI,” or in other words, expanding their own AI infrastructure to overcome reliance on nations who have their own data privacy and safeguard rules.
So despite the Trump administration’s attacks on Anthropic, developers are still raising funds and building at a frantic pace.
