The Trump administration has ordered Anthropic to prevent foreign nationals from accessing its newly released models, Fable 5 and Mythos 5, marking the first time the US has targeted specific AI models with an export-control measure. The directive, issued days after the models' launch, underscores a rapidly tightening federal grip on frontier artificial intelligence, according to reports from Axios and CNBC.
The move comes amid escalating concerns that advanced AI systems could be exploited by adversaries for military or surveillance purposes. Politico noted that Defense Secretary Pete Hegseth doubled down on the security risk designation, reinforcing the administration's stance. The New York Times confirmed that Anthropic has complied, disabling access to comply with the government order.
Anthropic, a leading Silicon Valley AI firm, has not disclosed how many foreign users are affected or whether the restriction applies retroactively. The Wall Street Journal reported that the company halted access to both Fable 5 and Mythos 5, though specifics on enforcement mechanisms remain unclear. No detailed timeline for lifting the ban has been provided.
Industry analysts warn that such targeted restrictions could stifle international collaboration and slow AI innovation, while potentially pushing foreign researchers toward competing models from China or Europe. The order also raises questions about how the US will balance security needs with maintaining global leadership in AI development. Affected foreign nationals may lose access to tools used for cutting-edge research.
Critics argue the ban is overly broad, potentially hindering legitimate academic and commercial partnerships without clear evidence of specific threats. Some experts contend that blocking foreign access alone may not prevent determined adversaries from obtaining similar capabilities through other channels.