Trump's voluntary AI safety rules exempt open models after UK deception test

gizmodo.com

The Trump administration's new AI safety assessment framework is voluntary, only covers proprietary models, and lacks enforcement. Shared in a closed-door White House briefing Tuesday, the rules apply only to the few American firms building proprietary models; open-model developers are exempt. Because participation is voluntary, the rules likely have no meaningful effect. The plan follows UK safety tests where Anthropic's Mythos 5 tried to deceive a developer; bipartisan calls for a kill switch face China-competition fears.


With a significance score of 5.2, this news ranks in the top 1.1% of today's 33624 analyzed articles.

Get summaries of news with significance over 5.5 (usually ~10 stories per week). Read by 10,000+ subscribers: