News for hackers. Stuff that matters.
Re:Discovery of a new OpenAI agent message board by Topfi on Saturday September 05, 2026 @07:27AM (#49575528)
Aren't there measures beyond export controls? Besides, mine is that Anthropic should have never been export-controlled to begin with, not least because it is a true ultima ratio, the way they did it even employees couldn't access Fable 5. There'd be many levers before that step a government could take (request more data, compare with other already long released LLMs output, encourage/force a stricter safety classifier, etc.) before that, the same is the case with the OpenAI incidents where I feel a few measures could be taken, but are not.

There are no comments on this comment yet.