How the US government is adopting AI: NASA reports 420 AI use cases in 2025, up from 18 in 2024, HHS reports 398, Energy 325, DOJ 295, Interior 234, and DHS 205
The White House is accelerating AI adoption across government, embedding the technology in policing, health care, defense and science.
Context & Ripple Effects
Federal AI adoption was already broad but uneven: a GAO review had identified more than 200 current agency applications and over 500 planned uses across nearly two dozen agencies. The new agency-level counts make the expansion visible in major operational departments rather than only in aggregate inventories.
This scaling follows the OMB's requirement for annual AI reporting and senior accountable officials, giving the White House a mechanism to turn experimentation into a more managed, comparable government-wide program.
First-order effects
- NASA, HHS, Energy, DOJ, Interior and DHS now have large reported AI-use portfolios, pushing AI oversight from isolated pilots toward routine management across science, health, energy and public-safety functions.
- Agency leaders must account for a much larger set of deployed uses under the federal reporting and ownership framework, especially where systems touch policing, health care or other high-impact decisions.
Second-order effects
- Shared reporting needs create pressure for agencies to standardize inventories, risk review and accountability practices, while technology vendors face more formalized federal buyer requirements.
- The concentration of use cases in HHS, Energy, DOJ and DHS raises the importance of governance approaches tailored to sensitive-domain deployments, rather than a single generic AI policy.
Third-order effects
- If agency portfolios continue to grow, federal AI procurement and operational controls could become a durable market-shaping layer for vendors serving regulated or mission-critical work.
- The central policy challenge shifts from encouraging adoption to governing a distributed estate of AI systems consistently across civilian, scientific and security-related agencies.
The trend: This is part of the industrialization of public-sector AI: centralized accountability rules are being used to scale AI deployment across agencies with very different missions.