City Detect in Stockton — AI Street Cameras That Flag Blight for an Understaffed Code-Enforcement Team

Stockton's code enforcement division, like many under-resourced city departments, could not keep up with reports of overgrown lawns, graffiti, peeling paint and other blight across a large city, relying on slow, complaint-driven inspections.
The city mounted cameras running City Detect's machine-learning software on municipal vehicles. As the vehicles drove their normal routes, the system analysed street-level imagery and assigned each location a relative 'blight score,' so code officers could prioritise where to send limited staff, rather than issuing automated penalties itself.
A five-day pilot in January 2024 identified roughly 4,000 code violations across more than 2,000 locations. Based on that data, Stockton's city council unanimously approved a $237,600-per-year contract, making Stockton the first California city to deploy the tool. In January 2025, the city expanded the approach into its RISE (Revitalizing and Improving Stockton through Education) programme, capturing street-level images at driving speeds up to 55 mph.
Officials and local reporting have framed the programme as 'education first,' using AI-flagged locations to prompt outreach and voluntary compliance rather than automatic citations, and enforcement decisions remain with human code officers. The performance figures come from the vendor-supported pilot and city reporting rather than an independent academic evaluation, and no public bias or accuracy audit of the underlying computer-vision model has been published.

Read the full analysis: https://www.cbsnews.com/sacramento/news/using-ai-to-bolster-enforcement-during-officer-shortage