Local government leaders across the U.S. are exploring the potential of artificial intelligence in the public sector.
Surveys show local government interest in AI is growing, particularly in the technology’s potential to increase efficiency. As AI ramps up, the policies surrounding city employees’ use of the burgeoning resource are still being written. Here are three cities with AI policies to watch.
San Francisco
San Francisco granted 30,000 city employees access to Microsoft 365 Copilot Chat last summer.
The city updated its AI policies around the same time, establishing five key guidelines. Those guidelines make human users accountable for anything created by AI that is used or shared; require secure AI tools; mandate thorough review of AI’s output; demand disclosure of AI use in public-facing or sensitive work; and ban “deepfakes” or the creation of fake images, audio or video that could be mistaken for real.
“It’s important that our workers are well-prepared to use new tools responsibly so that we are positioned to see benefits from emerging technologies well into the future,” City Administrator Carmen Chu said in announcing the Microsoft partnership.
Austin, Texas
Austin City Council last April passed an ethical AI use resolution that seeks to cement transparency and worker protections in the city’s use of AI.
The resolution requires an annual public audit of the city’s AI use as well as human review and oversight for all AI use. It also prohibits AI-based productivity scoring or behavioral monitoring without a human supervisor’s review and verification and creates a “no displacement without consultation” policy that blocks AI from significantly altering existing municipal job classifications without prior notice and consultation.
“This resolution ensures AI is used to support — not replace — public workers,” said Brydan Summers, president of Local 1624, which represents around 5,000 Austin and Travis County workers. “By requiring human oversight, banning continuous surveillance, and protecting workers from AI-only decisions, Item 55 puts the safety and dignity of the workforce first.”
The city in December put out a memorandum in compliance with the resolution. The state’s Responsible AI Governance Act also went into effect Jan. 1.
Seattle
Seattle in 2023 became one of the first cities in the nation to introduce an AI policy. After experimenting with nearly 40 different AI pilot programs, the city last fall unveiled updates to that policy.
The updated policy aims to further reduce bias in AI use and bars AI from emotion analysis and social scoring, as well as integration with weapons systems, and prohibits using AI with no human oversight. It also establishes AI training for city staff and a method for evaluating AI tools.
“Seattle remains focused on AI enabling better services and greater opportunity,” Rob Lloyd, chief technology officer for Seattle, said in a September news release. “Our second AI policy builds on lessons from almost 40 AI pilots and projects, while we upskill our teams, partner with leading companies, and apply our leading responsible use AI work.”
