US Artificial Intelligence Safety Agency Operates Without Leadership Amid White House Shift

The United States agency tasked with evaluating advanced artificial intelligence safety is operating without a permanent director and has halted the publication of its research findings. According to reports cited by the Associated Press, staff at the Center for Advancing Innovation and Standards for Super Intelligence (CAISSI) have recently been instructed to report informally to an office within the White House.

The leadership void follows repeated management turnover at the organisation, which operates under the Department of Commerce. Collin Burns resigned as director four days after his appointment in April following scrutiny over past ties to the AI developer Anthropic, and his successor, Chris Fall, stepped down three months later. Arvind Raman, head of the parent organisation National Institute of Standards and Technology, currently serves as acting director. Members of Congress submitted formal inquiries in August regarding the suspension of the agency's safety reports.

The shift toward White House oversight has drawn varying assessments from policy experts. Caroline Siegel Singh of the Federation of American Scientists stated that the centre was intended to rely on nonpartisan technocratic experts to guide policy in the public interest. Critics cited by the Associated Press expressed concern that direct executive oversight could lead to the softening of risk assessments regarding cyberattacks or bioweapons to accommodate the technology sector. Conversely, supporters of the shift argued that closer alignment with President Donald Trump could raise CAISSI's profile and secure additional funding beyond its current annual budget of $10 million.

The administration has increasingly framed its artificial intelligence strategy around national security rather than independent regulatory oversight. In September, President Trump called on technology companies to adhere to a voluntary self-policing agreement, though the administration did not outline verification mechanisms. When asked about CAISSI's operational status, White House representatives pointed to the newly established Super Intelligence Force, led by National Intelligence Director Jay Clayton. Amos Toh of the Brennan Center for Justice stated that treating artificial intelligence safety primarily through a national security lens risks marginalising consumer protections, privacy, and civil rights.

CAISSI's current resources remain smaller than those of its foreign counterparts. The agency's $10 million annual budget represents roughly one-eighth of the funding allocated to the United Kingdom's equivalent body. By comparison, the European Union's AI Office, established within the European Commission with enforcement powers since August, maintains a staff of more than 125 employees and is currently recruiting 40 additional personnel.

Written by Thomas Nussbaumer thomas.nussbaumer@alpineweekly.com