Four years after a wave of vehicle crime painted San Francisco as a lawlessness capital, data reveals a dramatic reversal. An analysis of San Francisco Police Department figures covering January through August 2026, compared to the same period in 2022, shows car thefts have decreased by approximately 59%, while car break-ins have plummeted by more than 82%.
City law enforcement officials, including District Attorney Brooke Jenkins, attribute this decline significantly to the expanded use of surveillance technology, particularly the AI-powered automated license plate reader (ALPR) network provided by Flock Safety. However, as controversies over privacy and potential misuse of the technology have led hundreds of other municipalities to suspend or terminate their contracts with Flock, San Francisco has chosen to stay the course, implementing stricter regulations instead.
The shift began after voters recalled DA Chesa Boudin in 2022, leading Mayor London Breed to appoint Jenkins, a vocal critic of Boudin’s lenient prosecution policies. Jenkins has since pursued a more aggressive legal strategy. Simultaneously, the SFPD overhauled its tactics, increasing patrols in tourist zones and deploying “bait cars” to target organized burglary rings.
In March 2024, voters approved Proposition E, easing restrictions on drone and camera usage. The SFPD promptly began installing Flock cameras in strategic locations. These devices capture images of passing vehicles, while AI software catalogs details such as make, model, color, and distinguishing features like roof racks or bumper stickers. This creates a searchable database that allows investigators to track vehicles even if they are not currently displaying license plates.
SFPD spokespersons described the ALPR system as instrumental in solving previously difficult cases, such as auto burglaries. Jenkins highlighted a specific incident where Flock cameras identified a vehicle linked to a prior day’s burglary. Officers then used a drone to surveil suspects in real-time as they smashed into a parked SUV, leading to arrests without the dangers associated with high-speed pursuits.
Jenkins argues that the visible presence of advanced surveillance also serves as a deterrent, convincing potential criminals that the risk of entering the city is not worth the potential consequences.
Despite these claims, empirical research on the specific impact of sprawling AI-driven license plate networks remains limited. A 2019 meta-analysis found that traditional CCTV cameras led to modest crime reductions, particularly in parking lots, where crime fell by roughly 37%. In contrast, studies on older, car-mounted plate readers generally found them ineffective at reducing overall crime, though they aided in vehicle recoveries.
A recent study by University of South Carolina researchers Ian Adams and Scott Mourtgos offers some of the most comprehensive evidence to date. Analyzing data from 216 law enforcement agencies, they found that deploying Flock cameras was associated with an 11% reduction in car thefts during the first year. The researchers utilized a “difference-in-differences” methodology to control for broader crime trends and conducted stress-tests comparing vehicle theft to shoplifting, finding that car theft declined more sharply.
Mourtgos noted that while the study has not yet undergone peer review and is not a randomized controlled trial, the results are suggestive. However, the researchers did not examine the impact on car break-ins, which have fallen even more drastically in San Francisco. Experts note that modern electronic immobilizers have made stealing newer cars harder, but they do not prevent thieves from smashing windows to steal belongings.
As other cities replace Flock with competitors like Axon, San Francisco’s approach—tightening rules rather than abandoning the technology—may serve as a test case for whether municipalities can harness high-tech surveillance to fight crime while protecting constitutional rights.
Leave a Reply