The Trump administration quietly disclosed a staggering 3,611 active or planned AI use cases across the federal government on 14 April, marking a 70% increase from the inventory published in the final year of the Biden administration. The list, released by the Office of Management and Budget (OMB), includes many plans to hand over sensitive governmental functions to AI, raising alarms about transparency and public accountability.
Disturbing AI Applications
Among the disclosed use cases are several that critics say evoke dystopian fiction. The Health and Human Services (HHS) office for children and families hired Palantir—a company known for its work with the military, CIA, and ICE—to scan grant applications for ideological alignment with the administration. The Federal Bureau of Prisons is developing an AI to assess misconduct risk for new inmates, potentially placing them in high-security confinement before any wrongdoing. The Department of Veterans Affairs is testing an AI that monitors crisis line calls to assess mental health and suicide risk. The Department of Energy is exploring autonomous AI control of nuclear reactors to respond to safety incidents.
Lack of Public Consultation
Despite the potential for responsible implementation, the inventory lacks essential details. Descriptions are often just a sentence, and public consultation is minimal. Only one of the cited examples—the Department of Justice—proposes public involvement. The OMB does not require public input for most use cases, as they are not classified as 'high impact.' The disclosure itself is obscure, appearing only on the federal CIO's GitHub account.
Potential Benefits and Risks
Experts note that some AI applications could be beneficial. For instance, Customs and Border Protection uses AI translation when human interpreters are unavailable. The number of such translation use cases rose from 58 under Biden to 70 under Trump. However, without meaningful public engagement, trust erodes.
International Examples
France's Digital Republic Act (2016) requires all government decision-making algorithms to be subject to public records requests, human appeal, and mandatory notification. Canada's 2025 AI use case registry mandates a transparent risk-scoring and impact assessment process, though it could be improved with public comment periods.
Call for Reform
Nathan E Sanders and Bruce Schneier, authors of 'Rewiring Democracy,' argue that the US should adopt algorithmic impact assessments and public comment processes at federal and state levels to ensure safe, equitable AI deployment in government.
