AI Policy Corner: How U.S. Cities Are Governing AI: Emerging Patterns and Collaborative Efforts

✍️By Yunzhe Liu

Yunzhe is a PhD Student in Political Science and a Graduate Affiliate at the Governance and Responsible AI Lab (GRAIL), Purdue University.

📌 Editor’s Note: This article is part of our AI Policy Corner series, a collaboration between the Montreal AI Ethics Institute (MAIEI) and the Governance and Responsible AI Lab (GRAIL) at Purdue University. The series provides concise insights into critical AI policy developments from the local to international levels, helping our readers stay informed about the evolving landscape of AI governance. This piece spotlights how U.S. cities are governing AI through a preliminary analysis of public policy documents, and explores how they are collaborating to address shared challenges.

While existing discussions of AI governance have largely focused on the federal and state levels, far less attention has been given to the local governments which are nonetheless emerging as key actors in this space. Cities across the U.S. serve as frontline implementers of AI tools with many local governments already adopting these technologies across a wide range of domains. This raises an important question: how are local governments governing the use of AI to harness its benefits safely, fairly, and responsibly?

To better understand the current landscape, a preliminary analysis of 24 publicly available local AI governance documents from major U.S. cities was conducted. It is important to note that this isn’t comprehensive, as some cities may not have publicly released their policies and smaller municipalities are likely underrepresented. Despite these limitations, the analysis provides useful insights into how local governments are designing AI governance.

What Patterns Emerge in Local AI Governance?

First, with the exception of New York City Local Law 144, which targets employers and employment agencies, most local AI governance documents focus on the use of AI within government operations. For example, the scope of Seattle’s AI Policy applies to “[a]ll staff (full-time, part-time), interns, consultants, vendors, contractors, partners, and volunteers who provide City services or otherwise act on behalf of the City,” highlighting a clear emphasis on internal governance.

Second, the vast majority of policy documents articulate guiding principles for the use of AI. These principles typically emphasize values such as transparency, accountability, fairness, and safety, providing a normative framework for how AI should be adopted and used by the government.

Despite such consensus, local governments differ in how they interpret these principles. For instance, the City of Spokane places greater emphasis on the technical transparency of AI systems, defining transparency as “[a]n AI system, its data sources, operational model, and policies that govern its use are understandable and documented.” By contrast, the City of San José extends the concept to include public-facing transparency, emphasizing that “[t]he purpose and use of systems is proactively communicated and disclosed to the public.” In practice, the City of San José also maintains a publicly accessible AI inventory that documents the functions and technical details of its used tools.

Third, 7 out of the 24 policies explicitly identify the prohibited use of AI, but the level of specificity varies considerably. For example, the Metropolitan Government of Nashville and Davidson County broadly states that “fully automated decisions that do not require any meaningful human oversight and that may substantially impact residents” are prohibited. However, some other cities provide more detailed lists of prohibited AI applications. San José, for instance, prohibits a range of high-risk use cases, including emotion analysis, social scoring, and cognitive behavioral manipulation of people. Interestingly, the City of Long Beach goes even further by naming specific AI tools and companies, such as Deepseek, Otter.ai and Read AI, due to their “serious concerns of transparency, data privacy and security.”

How Are Cities Collaborating to Govern AI?

In addition to individual efforts, cities are increasingly collaborating with one another to address shared challenges in local AI governance. In particular, the GovAI Coalition, launched by the City of San José and composed of government agency members, provides a set of policy templates and practical resources to help agencies jumpstart responsible AI practices. A more regional example can be found in Michigan, where the Science, Technology, and Public Policy (STPP) program at the University of Michigan and the Michigan Municipal League jointly released the Artificial Intelligence Handbook for Local Government, offering guidance and resources to help municipalities use AI technologies safely and effectively.

Further Readings

Image link: https://unsplash.com/photos/a-view-of-a-city-at-night-with-the-moon-in-the-sky-GA0OGHrNDHg