US AI Safety Institute Shifts Informal Oversight to White House

The federal AI safety testing organization has begun reporting informally to a White House office while operating without a permanent leader. The agency evaluates potential national security and technological risks in frontier AI developments.