President Donald Trump signed an executive order establishing a voluntary framework for federal vetting of advanced frontier AI models before public release. The directive gives government agencies a 30-day testing window to assess national security and cybersecurity risks. Participation is optional for AI developers to avoid hindering innovation and US competitiveness against rivals like China. The order follows concerns over models such as Anthropic's Claude Mythos, which demonstrated advanced vulnerability discovery capabilities.
Tracked by ByteBrief