PRESIDENT Donald Trump signed an executive order establishing a voluntary framework for federal assessment of advanced AI models' national security risks prior to public release. This order allows a 30-day testing period for government agencies to evaluate cybersecurity risks, encouraging participation from AI developers while aiming not to stifle innovation.
Industry experts express mixed reactions on the effectiveness of voluntary compliance, suggesting issues such as regulatory capture and the need for more robust governance frameworks. Feedback highlights concerns about AI agents' behavior once deployed and the necessity for improved vulnerability management and automation. Overall, the executive order is seen as a step towards addressing security risks associated with AI, but challenges in implementation and industry responses remain significant.