New research from Gallup, published in March 2026 by researcher Christos Makridis, shows AI adoption in the U.S. public sector has surged to 43% of employees using AI at least occasionally in Q4 2025, up from 17% in Q2 2023 and 28% in Q2 2024. That figure now exceeds the private sector's 41% — a reversal that would have seemed unlikely two years ago given the stricter governance frameworks, procurement requirements, and cautious risk culture that historically shaped federal technology adoption. Among public-sector workers, 21% report using AI daily or multiple times per week. A year ago, most of those users probably weren't.
Manager behavior matters more than tool access. In public-sector agencies where managers actively support AI experimentation, 65% of employees are frequent AI users. In low-support environments, that drops to 37% — a 28-percentage-point gap that dwarfs any difference attributable to budget or software availability. The pattern mirrors private-sector adoption research and points to the same conclusion: whether an organization <a href="/news/2026-03-14-longitudinal-study-ai-tools-boost-developer-productivity-10-percent-not-hyped-2-3x">captures productivity gains</a> depends less on what tools it buys and more on whether managers actually encourage staff to use them.
A strategic gap runs alongside the adoption surge. Only 37% of public-sector organizations report having a clear AI strategy, compared to 53% of private-sector counterparts. Gallup attributes much of the bottom-up adoption to the low barrier to entry of consumer-grade generative AI tools, which employees can adopt without formal IT procurement or training. <a href="/news/2026-03-14-amazon-employees-say-ai-is-just-increasing-workload-study-confirms">Widespread use without formal governance</a> carries real compliance risks for agencies handling sensitive data — many popular AI tools have never been evaluated against FedRAMP, FISMA, or NIST 800-53 requirements. The Government Accountability Office flagged a severe shortage of digital expertise in government as recently as 2023, and Lightcast data shows AI-related job postings represent less than 0.3% of public-sector listings. The agencies with the fastest adoption growth are often the least equipped to audit it.
On the policy front, Memorandum M-25-21 signals a deliberate shift in federal AI governance away from risk management and toward broader agency-level experimentation, which could push adoption rates higher still through 2026. The practical implication for vendors: federal AI procurement is moving from scattered pilot programs toward enterprise-scale contracts, and the agencies most likely to buy at scale are the ones already running unsanctioned tools across their workforce.