Anthropic Employees Accuse Trump Administration of Targeting Them
TL;DR
The Trump administration reportedly used export controls to force Anthropic to restrict access to its newest models, Fable 5 and Mythos 5, for foreign nationals, including people inside the US. Anthropic then disabled broader access because enforcing the order by nationality was not realistic on short notice. The restriction would also affect some of its own employees.
Nauti's Take
Anthropic spent years warning that powerful models can be dangerous. Now the company is getting a hard version of that logic applied to itself.
The problem is not that security questions are being asked, but how opaque and ad hoc the intervention looks. If a government can effectively freeze a frontier model with only hours of notice, the industry needs clear criteria instead of case-by-case power politics.
Briefingshow
This is bigger than one company fight: the US is testing whether advanced AI models can be treated like strategic export goods. If access depends on nationality, security suspicion and political trust, the global AI market becomes far less predictable. For companies, choosing an AI provider now also carries geopolitical risk.
Sources