Staff at the US center for AI safety are reporting to the White House, raising independence concerns

A small U.S. government unit tasked with assessing the safety of new AI systems, the Center for Advancing Innovation and Standards for Super Intelligence (CAISSI), has reportedly begun informal reporting to an office in the White House, raising concerns about independence. Two former officials told The Associated Press this change was discussed, while CAISSI has also lacked a permanent leader and stopped publishing its research. Critics worry executive involvement could politicize decisions affecting areas like cybersecurity and biosecurity, while supporters say closer alignment with President Donald Trump could bring more prominence and funding. CAISSI receives $10 million per year, and its leadership turnover has been frequent since 2024.