This paper analyzes the composition of federal AI adoption using the Office of Management and Budget's 2024 and 2025 consolidated AI use case inventories. Focusing on the principal federal statistical agencies overseen by the Interagency Council on Statistical Policy (ICSP), it finds that reported growth is concentrated in developer productivity tools and enterprise copilots rather than in production statistical systems. A composition decomposition separates real portfolio shifts from arithmetic dilution caused by a doubling of the inventory. The Buy track shows a 46% concentration in a single vendor's products, driven by the timing of FedRAMP authorizations rather than a deliberate portfolio choice. The paper positions federal statistical AI adoption relative to the commercial frontier and identifies structural constraints that shape the gap. The views expressed are the author's own and do not represent the U.S. Census Bureau or the Department of Commerce.
Brock Webb (Sun,) studied this question.