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

As concerns over uncontrolled artificial intelligence agents rise, the U.S. government division responsible for assessing AI safety is facing independence issues. Known as the Center for Advancing Innovation and Standards for Super Intelligence (CAISSI), a small division within the Commerce Department, it has lacked a permanent leader and ceased publishing research.

Recently, officials from CAISSI were asked to informally report to the White House, raising questions about potential politicization of AI safety decisions. This shift could impact how the center's findings are received, possibly influenced by the tech industry, which largely polices itself. Caroline Siegel Singh, an AI program manager, argues that nonpartisan experts are needed to best serve the American public.

The center, funded at $10 million annually, has seen high turnover among its leaders, with three directors leaving in quick succession since appointing a new one in April. In August, Congress began questioning why CAISSI halted public releases of its AI safety reports. While the White House did not comment on CAISSI specifically, it acknowledged the establishment of President Trump's Super Intelligence Force and his commitment to "continued U.S. leadership on SI for the benefit of the American people."