> When anthropic contractually stopped the US government from using the platform for certain purposes that they could not prevent the average user from pursuing they created a de facto national security risk.

Anthropic had two red lines: 1) no lethal autonomous warfare without a human in the loop, and 2) no mass surveillance of Americans. How is that something the average user is capable of? How are those red lines creating a national security risk?

These were not special terms for the U.S. government, they were a part of the usage policy since 2024. Governments are not in the same class as the average user, their capability and potential for misuse of the service is not comparable. There is no reason why the U.S. government should dispute those terms, and their reaction was dystopian.