2026 Spring Seminar Series Seminar: Catherine Annis University of Kentucky

Institutions and Innovation: Examining Reported AI Adoption in Federal Agencies

Abstract: The adoption of artificial intelligence (AI) across federal agencies represents a critical inflection point in public sector innovation. Yet, the determinants of reported AI adoption in public organizations remain poorly understood. This study applies the technology-organization-environment (TOE) framework to examine how agency-level characteristics shape reported AI adoption in the U.S. federal government. Leveraging data from the 2024 Federal AI Use Case Inventory and September 2024 FedScope Workforce Dataset, we analyze 470 agencies mandated to report AI use cases under Executive Order 13960 using ordinary least squares and logistic regression. The findings highlight relevant contextual variables associated with reported early adoption, offering new insight into AI governance within federal bureaucracies.