San Francisco should end Flock Safety license plate reader program
## CONTEXT
San Francisco, like many urban centers, has grappled with balancing public safety and civil liberties. Since 2022, the San Francisco Police Department (SFPD) has deployed Flock Safety’s automated license plate recognition (ALPR) cameras at key intersections and high-crime areas. These cameras capture time, date, GPS location, and license plate of every passing vehicle, storing data for up to 30 days. While touted as a tool for recovering stolen vehicles and solving serious crimes, the system has drawn sharp criticism from privacy advocates who argue it enables indiscriminate mass surveillance. In July 2026, the Los Angeles Police Department terminated its Flock Safety agreement, citing privacy concerns and low success rates in court. This decision by a fellow large California city creates a pivotal moment for San Francisco. The question is not whether the technology works in isolation, but whether its societal costs—chilling effects on free movement, potential for misuse, and data vulnerability—outweigh its investigative benefits. Other cities, including Oakland and Portland, have similarly reconsidered or restricted ALPR use. The current situation demands a careful re-evaluation of San Francisco’s commitment to Flock Safety.
## PROBLEM
The core problem is that Flock Safety cameras operate as a blanket surveillance tool, capturing data on all vehicles—not just those connected to crimes—without a warrant or individualized suspicion. This practice erodes Fourth Amendment protections and disproportionately affects marginalized communities who already face heightened police scrutiny. A 2023 EFF analysis of Flock data practices revealed that police departments can access the data of any driver who passes a camera, often without judicial oversight. The harm is twofold: first, innocent citizens are tracked without consent, creating a chilling effect on political assembly, travel to sensitive locations (e.g., reproductive health clinics), and everyday activities. Second, the data is vulnerable to breaches or internal misuse; in 2024, a leak of Flock data from a Midwest police department exposed the movement patterns of over 200,000 residents. The cost of inaction is continued erosion of privacy trust. San Francisco already has a history of surveillance technology controversies—from stingrays to face recognition bans—and maintaining Flock undermines the city's progressive reputation. Moreover, the financial cost is significant: SF’s contract is estimated at $1.5 million annually, with questionable public safety returns. LAPD found that Flock cameras contributed to only 6% of stolen vehicle recoveries and were rarely used as primary evidence in prosecutions. Without action, San Francisco risks perpetuating a system that collects vast data with minimal accountability.
## PROPOSED SOLUTION
The solution is for the San Francisco Board of Supervisors to pass an ordinance immediately terminating the SFPD’s Flock Safety contract and mandating the removal of all cameras within 90 days. Alternative approaches considered include implementing strict data retention limits (e.g., 48 hours) and requiring warrants for any query. However, these half-measures fail to address the fundamental issue of mass collection—once data is captured, the potential for abuse remains, and oversight burdens fall on already strained privacy offices. The preferred action is a clean cut, modeled on LAPD’s recent termination. The SPADE framework guides this: Situation—existing contract expires in 2027 with a 6-month cancellation clause. Decision—end the agreement immediately. Action—the city attorney gives notice to Flock, cameras are physically removed, and data is deleted. Process—a moratorium on any future ALPR contracts unless a voter-approved ballot measure explicitly authorizes them. Execution—the Police Commission oversees removal, and the Controller’s Office audits Flock to ensure no lingering data. To address the loss of a crime-solving tool, the SFPD will allocate the saved funds to community policing initiatives and victim services, and will rely on existing resources like gunshot detection (ShotSpotter, though controversial) and investigative detective work. A civil liberties oversight committee will be established to review any future surveillance technology proposals. This solution aligns with San Francisco’s 2019 surveillance technology ordinance that required public approval for new tools—Flock was previously grandfathered in, but a fresh ordinance would correct that oversight.
## EXPECTED IMPACT
The immediate benefit is the restoration of privacy for every person driving through San Francisco. Approximately 2 million vehicles cross camera-equipped intersections annually; their movement data will no longer be stored in a private vendor’s cloud, vulnerable to warrantless police queries or breach. Community trust in local government is likely to increase, especially among advocacy groups that have long opposed ALPRs. Metrics of success include a measurable reduction in community complaints about surveillance (target: 50% decrease in such complaints to the Office of Civil Rights within one year). However, there is a potential trade-off: without Flock, the recovery rate of stolen vehicles may drop from the current 78% (which includes all methods) by an estimated 5-10 percentage points. Yet comparable jurisdictions that ended ALPR programs, such as Berkeley, saw no significant change in auto theft clearance rates because stolen vehicle identification still relies on owner reports and CHP databases. The financial impact is positive: savings of $1.5 million annually can be redirected to overtaxed city programs. The broader societal impact is symbolic: San Francisco would reassert itself as a leader in privacy protections, encouraging other California cities to follow. Downside risk: some law enforcement resistance, but this can be mitigated through transparent working groups demonstrating that traditional investigative techniques remain effective. Overall, the proposal strengthens democratic values without sacrificing core public safety functions.
## DECISION LENS
| | If this passes | If this doesn't pass |
| --- | --- | --- |
| What will happen | Flock cameras removed; surveillance data deleted; privacy restored; savings reallocated; trust rebuilt. | Continuation of mass license plate tracking; potential privacy litigation; erosion of public trust; $1.5M spent annually. |
| What won't happen | Losing all stolen-vehicle recovery capability; cameras remain as operational tools. | A robust debate on surveillance; immediate privacy protection; policy alignment with LA. |
## PRECEDENTS
EXAMPLE: Los Angeles, CA — What: LAPD ended its Flock Safety agreement citing privacy concerns and low court usage. The decision followed a city council vote mandating contract review. — Outcome: Removal of approximately 300 cameras citywide; data deletion; no reported increase in crime-solving delays. — Outcome: Removal of approximately 300 cameras citywide; data deletion; no reported increase in crime-solving delays.
EXAMPLE: Oakland, CA — What: Oakland Police Department limited use of its ALPR system (not Flock specifically) to a 48-hour data retention policy after public outcry. — Outcome: Reduced potential for profiling; community groups reported increased satisfaction; vehicle theft recovery rates remained stable. — Outcome: Reduced potential for profiling; community groups reported increased satisfaction; vehicle theft recovery rates remained stable.
EXAMPLE: United Kingdom — What: Several UK boroughs decommissioned ALPR networks after audits showed they were ineffective in reducing crime and raised disproportionate privacy concerns. — Outcome: Removal of 40% of cameras in pilot areas; public survey showed 70% support for the removal; crime rates unchanged. — Outcome: Removal of 40% of cameras in pilot areas; public survey showed 70% support for the removal; crime rates unchanged.
July 13, 2026
