US Government Revealed 3,611 AI Use Cases in Government

The Trump Administration has declassified data on the large-scale use of artificial intelligence in federal authorities: 3,611 active or planned cases — 70% more than under the previous administration.

Silent Recognition

On April 14, the Office of Management and Budget (OMB) published an updated registry of AI applications in U.S. federal agencies.

Notably, this occurred without press conferences or public hearings — the data appeared as a technical commit to the government repository on GitHub. The list grew by 70% compared to the analogous registry from the final year of Biden's presidency and covers thousands of scenarios in which algorithms make or support decisions on behalf of the state. For scale: just a few years ago, there were hundreds of pilot projects, primarily in document management.

Today the figure has surpassed three and a half thousand. Some of them are relatively harmless auxiliary tasks: automating request processing, database searching, document classification. But a significant portion of the registry concerns decisions that directly affect the lives and freedom of specific people.

What Comes Under Algorithmic Control

The registry covers a wide and in some ways alarming spectrum of sensitive state functions. Among the tasks delegated to AI:

• Risk assessment systems in legal proceedings — influence decisions on arrest, bail, and parole
• Managing access to state services in healthcare and social welfare
• Monitoring nuclear reactor safety and critical infrastructure
• Border control and immigration decisions
• Predictive analytics in law enforcement

These are not abstract prospects — systems are already being deployed or are planned for launch across all levels of federal administration.

Transparency That Doesn't Exist

Nathan E. Sanders from Harvard's Berkman Klein Center and Bruce Schneier from the Kennedy School of Government — authors of the book "Rewiring Democracy: How AI Transforms Our Politics, Government, and Citizenship" — formulate the core problem as follows:

"This is a transfer of decision-making processes from human to machine

at an enormous scale — on matters of personal freedom, health and well-being of people, nuclear facility safety, and much more."

The key question is not the fact of AI application itself, but the absence of accountability and appeal mechanisms. When an algorithm makes a decision on immigration status or the degree of risk for a prisoner's release — who is responsible for error? How do you contest a decision if the criteria for making it are not disclosed? Many state algorithms operate like "black boxes" — their internal logic is closed even to regulators.

What This Means

The disclosure of 3,611 cases is a step toward transparency, but the manner in which it occurred is telling. The public learned not through Congressional hearings and not through official statements, but through a technical commit on GitHub. When a large-scale transfer of state functions to algorithms occurs without public discussion, the risk of systemic errors grows, and the tools to correct them — narrow. The question is no longer whether AI will manage critical state functions. It already does. The question is who controls the AI itself.