Resource Automation + AI Mitigating Harm + Bias

Legacy Procurement Practices Shape How U.S. Cities Govern AI: Understanding Government Employees’ Practices, Challenges, and Needs

This paper explores how legacy procurement processes in U.S. cities shape the acquisition and governance of AI tools, based on interviews with local government employees.

Organizations: Carnegie Mellon University, University of Pittsburgh Institute for Cyber Law, Policy, and Security, TechBetter

Authors: Nari Johnson, Elise Silva, Harrison Leon, Motahhare Eslami, Beth Schwanke, Ravit Dotan, Hoda Heidari

Published Year: 2025

View Source

Download Resource

The study examines how U.S. cities acquire AI technologies through legacy procurement systems originally designed for conventional goods and services.

Based on interviews with 19 city officials from seven cities, the paper identifies how these practices influence decision-making power, risk mitigation, vendor relationships, and public accountability. It outlines cities’ emerging efforts—such as AI-specific reviews and contract language—to govern AI more ethically, while highlighting ongoing challenges like limited vendor transparency, uneven oversight, and the need for shared governance responsibilities. The authors call for more empirical and community-informed support to help cities mitigate AI harms during procurement.