SCOOP: Trump AI Executive Order Will Likely Punt on Security Concerns

President Donald Trump is expected to sign an executive order on artificial intelligence safety later this week, though the directive will avoid addressing critical national security risks posed by new AI models. The order will assign the Office of the National Cyber Director a 60-day deadline to develop a cybersecurity framework, focusing on voluntary vetting of AI labs’ pre-deployment models. However, it does not specify actions to take if a model is deemed dangerous during vetting, leaving key questions unanswered. The draft order was influenced by a prior proposal requiring government officials to review AI models before public release, which had broad voter support—82% according to a YouGov/Institute for Family Studies poll. Some officials pushed for mandatory pre-deployment reviews for labs working on government contracts, but the final version adopts a voluntary approach. The Center for AI Safety and Innovation (CAISI) within the Commerce Department already offers such vetting, though details of recent agreements with Google DeepMind, Microsoft, and xAI were later removed from the department’s webpage. The order will also task agencies with hiring more personnel to strengthen cyber preparedness in critical infrastructure. Leadership of the Office of the National Cyber Director, Sean Cairncross, has faced scrutiny over technical qualifications, as he lacks a background in cybersecurity despite prior political roles, including as chief operating officer of the Republican National Committee. A White House official defended Cairncross, stating his office collaborates closely with industry leaders like OpenAI and Microsoft to address AI-related cybersecurity challenges. The executive order will involve the Office of the National Cyber Director and intelligence community members in the vetting process, though its voluntary nature and delayed risk-mitigation measures have sparked internal debates. The White House has emphasized that any policy will stem directly from the president’s office, with ongoing engagement across government and industry to protect national security. Industry leaders, including Jason Kwon of OpenAI and Brad Smith of Microsoft, have reportedly praised Cairncross’s efforts in advancing cybersecurity objectives.