U.S. AI Safety Center Now Informally Reporting To White House

Staff at the federal Center for Advancing Innovation and Standards for Super Intelligence have in recent weeks begun informally reporting to a White House office instead of operating at arm's length inside the Commerce Department, according to two former officials familiar with the shift, raising fresh concerns about the center's independence as it evaluates fast-moving artificial intelligence systems.^[1]

The reporting line change, which has not been publicly announced, means staff at the AI safety center, known as CAISSI, now take informal direction from the White House even though the office was set up as an independent technical body with its own $10 million annual budget to scrutinize powerful AI models.^[1] The move comes as President Donald Trump expands his direct role in AI policy, creating a new "Super Intelligence Force" and ordering his Council of Advisors on Science and Technology to prepare a formal assessment of AI risks, according to a White House statement.^[1]

CAISSI was established within the Commerce Department as an independent division specifically to test advanced AI systems, publish safety benchmarks, and advise regulators and industry without day-to-day political oversight, an arrangement supporters said was crucial to winning public trust in any warnings or recommendations it might issue.^[1] The center has struggled to maintain stable leadership, however: its second director resigned after only a few months earlier in 2026, and the head of its parent standards institute has been serving as acting director since then, leaving a vacuum that former officials say made it easier for the White House to assert informal control.^[1]

Concerns about transparency around CAISSI's work predated the new reporting arrangement. The center abruptly stopped publishing its AI safety and performance reports earlier in 2026, halting a stream of public evaluations that companies, researchers and lawmakers had begun to rely on to understand how the latest systems behaved under stress tests.^[1] That pause prompted pointed questions from members of Congress at an August oversight hearing, where lawmakers pressed Commerce and White House officials on whether politically sensitive findings about leading AI companies were being delayed or buried.^[1]

Supporters of stronger presidential control say closer coordination between CAISSI and the White House could ensure that technical findings feed directly into national security planning, including the administration's new Super Intelligence Force and the upcoming risk report from the Council of Advisors on Science and Technology.^[1] But former officials who spoke about the informal reporting shift argue that routing an ostensibly independent standards body through the president's political staff undermines its credibility at the exact moment when its assessments could shape regulation of some of the most powerful technologies in decades.^[1] They warn that, without clear statutory protections and transparent publication of its findings, CAISSI risks becoming another policy arm of the executive branch rather than the neutral AI safety referee it was billed to be.

1. PBS News