Generative AI Use in Public Administration: Early Adoption and Governance Controls

Abstract

Background: The expansion of policy activity around generative AI has exposed a potential gap between formal design and day-to-day implementation, particularly in relation to public administration.

Objective: We evaluated the administrative operation of policy concerning generative AI, emphasizing transparency, practical implementation conditions, and public administration.

Methods: A cross-sectional survey of 358 relevant officials, professionals, service users, or regulated actors measured implementation quality, institutional capacity, procedural clarity, transparency, and perceived effectiveness. Descriptive statistics and multivariable regression were used to examine factors associated with stronger implementation.

Results: The final analytic sample for generative AI included 353 respondents. Consistent implementation was reported by 41%, while 17% identified significant capacity or procedural barriers. Organizations with clearer responsibilities, more capable staff, and accessible formal guidance recorded better adjusted implementation outcomes.

Conclusion: The study suggests that variation in outcomes related to generative AI is driven substantially by how institutions organize implementation and oversight. This means that AI governance in the public sector needs enforceable controls, capable staff, documented accountability, and meaningful human review.

Study design: multi-agency survey.

Publication history: Originally published in print on 15 June 2023, Volume 3, Article 1. Online abstract record added on 16 September 2026. Full-text PDF will be added when the digital galley is ready.