Governing Artificial Intelligence For Municipal Services - State of The Art, Patterns, and Practical Evidence

This paper synthesise the state of the art of Artificial Intelligence (AI) in munic-ipal Public Administration, explain when and how to adopt Large Language Models based assistants with governance, compare rule-based Q&A chatbots to AI chatbots using evidence from international deployments, and map municipal domains to representative data and AI patterns. We also summarise European Union (EU) data/security frameworks relevant to municipalities and propose a reference approach for continuous assurance. The contribution is a practical, ev-idence-based guide that links technology choices (e.g., Retrieval-Augmented Generation, tool calling) with public-sector controls (EU AI Act, United States AI Risk Management Framework) and measurable outcomes, supporting mu-nicipalities to move from pilots to reliable, scalable services.