Chula Vista's city council voted unanimously this week to adopt a sweeping new policy governing how the city buys, deploys and oversees artificial intelligence, barring departments from using real-time biometric surveillance or predictive policing tools without council authorization. The policy, approved Tuesday, applies to every city department, contractor and partner agency, and requires human review whenever an AI-assisted decision affects individuals or communities.
The vote, reported by the San Diego Union-Tribune, caps months of work by city staff and the Chula Vista Privacy Protection and Technology Advisory Commission, which reviewed the draft policy in September but took no formal action on it. Mayor John McCann made the motion to adopt the policy, saying it contains strong protections for individual privacy and personal information. Deputy Mayor Cesar Fernandez called it a very strong start, though he urged the city's forthcoming AI working group to consider how its protections could avoid becoming obsolete or administratively weakened as the underlying technology grows more powerful.
What the Policy Actually Prohibits
Under the new rules, city departments must evaluate AI tools for bias, disparate impacts and workforce implications before deployment, and the policy prohibits fully automated decisions made without meaningful human oversight. It also bars covert behavioral tracking and blocks employees from using free public generative AI platforms that require them to create individual accounts. Low-risk tools — brainstorming, proofreading and grammar-correction software not integrated into city systems or used with confidential or personally identifiable information — are exempt from the stricter requirements.
City staff must also build a centralized, publicly disclosed inventory of AI systems in use, and any system that undergoes a material change after procurement, such as added features or capabilities, triggers a new review before it can keep operating. David Graham, the city's economic development director, will serve as Chula Vista's AI coordinator during the initial implementation phase. Graham said the policy represents a safe and responsible framework that recognizes both city and community values. Adrianna Relph, the city's special projects and legislative manager, said the policy does not delegate decision making to AI.
A City Built Around Testing Public-Safety Tech
The policy lands in a city with an unusually long history of testing automated public-safety technology. Chula Vista became the first U.S. city authorized by the Federal Aviation Administration to fly drones beyond visual line of sight for 911 emergency responses, launching its Drone as a First Responder program in 2018, according to inewsource. By mid-2026, those drones had logged more than 25,000 emergency missions, resolving roughly a quarter of calls without needing to send patrol officers to the scene, per the Electronic Frontier Foundation.
That history helped drive the creation of the oversight body now reviewing the city's AI policy. Chula Vista established its Privacy Protection and Technology Advisory Commission in July 2023 under municipal code, following recommendations from a 12-member resident and expert task force formed the year before. The commission discussed independent certification and third-party auditing of AI systems, enforcement of the new policy, employee training requirements and potential workforce impacts, with its recommendations folded into the final version the council adopted.
Drone Deal Pause Set the Stage
The policy's arrival follows a separate, closely watched episode just weeks earlier. In September, Chula Vista police paused a proposed deal to acquire automated drones from surveillance vendor Flock Safety after presenting the plan to the same privacy commission and facing public pushback over cost and privacy, a move Hoodline covered in detail. That pause came not long after the department had presented plans, in December 2025, to shift the city's Real-Time Crime Center software onto Axon's Fusus platform under a five-year, $677,692 contract that would consolidate live drone feeds, automated license plate readers and municipal video streams.
Those episodes illustrate the balancing act the new policy is meant to formalize: giving the city room to keep testing public-safety technology while building in the kind of transparency and human oversight that residents have demanded. The policy itself carries no fiscal impact from adoption, but it does require a meet-and-confer process with city labor groups before it takes effect, a condition the council added to the motion at staff's request.
Next Steps and a Regulatory Gap It Fills
Chula Vista now has 120 days to convene its AI working group, designate officials and AI leads in each department, inventory AI systems already in use, and finalize an AI handbook. The working group must revisit the policy at least every two years, and city staff are expected to report back to the privacy commission with an implementation update one year after adoption.
The move also reflects a broader pattern across California, where state-level efforts to set AI and biometric-surveillance guardrails have repeatedly stalled. Assembly Bill 1814, which would have regulated law enforcement facial recognition statewide, died in the state Senate Appropriations Committee in August 2024 after civil rights groups and wrongful-arrest survivors argued its protections were too weak, according to the ACLU. Governor Gavin Newsom separately vetoed Senate Bill 1047, a broader frontier AI regulation bill, in September 2024. With state action stalled, cities like Chula Vista and San Jose, which launched its own public AI system inventory back in January 2023, have increasingly been left to write the rules themselves.
