It’s also punitive - Anthropic can’t comply, it locks their research staff which is, like any frontier lab, mostly international out.

The objective isn’t national security because we already know how that goes https://www.bbc.com/news/stories-54695598

There’s no path to compliance and the decision is arbitrary - the model capabilities are not officially assessed by any visible criteria and it prevents export of models based on these non criteria forward.

All days before IPO.