Westchester ALPR Surveillance Lawsuit Ledger: 1.6 Billion Plate Reads, 575 Cameras, And The Missing Guardrails

The Short Version

A new class-action complaint in New York turns Westchester County's automatic license-plate-reader network into a clean accountability test: when does public-safety technology become a countywide movement file on ordinary drivers, and who signed off on that trade?

The lawsuit was filed June 9, 2026, in Westchester County Supreme Court by four named plaintiffs represented by the Policing Project at NYU Law, the Knight First Amendment Institute, the New York Civil Liberties Union, and Freshfields. The defendants are Westchester County, the county Department of Public Safety, Commissioner Terrance Raynor, and Chief James Luciano. The complaint challenges the Westchester County Police Department's vehicle-surveillance system as unconstitutional under the New York State Constitution and as beyond the department's statutory authority.

That is the allegation, not a final court finding. The complaint itself carries the standard New York e-filing caution that the document had not yet been reviewed and approved by the County Clerk at the time of the PDF printout. The Associated Press reported the same day that a county spokesperson said Westchester had not yet received or reviewed the referenced lawsuit. That response matters. BadPD should not treat a complaint as a verdict.

But complaints can still be public records worth reading, especially when they attach numbers to a surveillance system that most drivers never knowingly opted into. Here, the plaintiffs allege a system with at least 575 automatic license plate readers, a searchable database of roughly 1.6 billion vehicle records, a default retention period of at least two years, and direct or networked access available to dozens of federal, state, and local law-enforcement users.

The BadPD angle is not "never use cameras." The accountability question is narrower and harder to dodge: if a county police department wants to turn public roads into a long-term searchable map of people's movements, where are the statute, budget trail, retention rule, audit log, access list, warrant standard, public reporting, and independent oversight?

What The Complaint Says The System Does

The complaint describes Westchester's vehicle-surveillance system as three parts: hundreds of automatic license plate reader cameras, a database that stores the scans, and analytic software powered by tools supplied by Rekor Systems. The cameras allegedly record passing vehicles around the clock and capture license plate numbers, vehicle make and model, color, location, timestamp, direction of travel, and visual identifiers such as bumper stickers or decals. The complaint also says the system can capture photographs and video that may include drivers and passengers.

The most important numbers are not subtle. The plaintiffs allege that Westchester County Police Department had at least 575 ALPRs connected to the system by the time of the complaint, compared with 36 when the system began in 2013. They allege the system collected 264,303,687 vehicle reads in 2024 alone, more than five million reads per week. They also allege that the department estimated in 2023 that the system contained roughly 1.6 billion reads.

The accountability problem is the denominator. According to the complaint, only 0.8 percent of the 2024 reads appeared on a hotlist, meaning more than 99 percent of the collected records were not tied to a suspected crime or investigation when gathered. If those figures are proved, Westchester's system is not just catching stolen cars or wanted vehicles. It is collecting the movements of nearly everyone else as the normal operating condition of the network.

The complaint alleges that WCPD stores all vehicle-surveillance data in a searchable database for at least two years by default. It also alleges that some named plaintiffs' data were retained for even longer, with no clear notice or explanation. That is a major fact to verify because retention is where a quick camera hit becomes a retrospective movement dossier. A one-day hotlist alert is one kind of tool. A two-year searchable travel history is another.

The system allegedly operates out of Westchester's Real Time Crime Center, which the complaint says gives officers across the county, including nearly 40 local police departments, access to current and historical information derived from the ALPR system. The complaint says WCPD also shares direct access with at least 55 federal, state, and local law-enforcement agencies, including ICE, the FBI, and the DEA. It further alleges that WCPD participates in Rekor's Public Safety Network, which can make real-time and historical vehicle-surveillance data available to other participating agencies.

That sharing claim is where the story stops being just a county privacy case. If local data can be searched by outside agencies, then a local camera system can become a national workaround. Residents may be told one thing by local sanctuary, privacy, or public-safety policies while the data layer quietly gives outside users another path.

What Is Confirmed, Alleged, And Still Pending

Confirmed: the Policing Project published a June 9, 2026 release announcing the case and linking to the complaint. The complaint names Westchester County, the county Department of Public Safety, Commissioner Terrance Raynor, and Chief James Luciano as defendants. The complaint says the plaintiffs seek a declaration that the system is unlawful and an order prohibiting WCPD from operating it. AP independently reported the filing and the main claims, including the 1.6 billion record figure, the nearly 600-camera description, and sharing with more than 50 outside agencies.

Alleged: the court has not decided that WCPD violated the New York State Constitution, exceeded its authority, improperly shared data, or operated without adequate safeguards. Those are claims in a civil complaint. The plaintiffs allege the system lacks meaningful rules governing collection, retention, searches, analysis, and sharing. They allege officers do not need a judicial warrant to search the database or run analyses. They allege the camera placement disproportionately burdens Black and Latine neighborhoods near the Westchester-Bronx border. Those claims deserve careful treatment because they may be contested.

Pending: the county's answer, any motion to dismiss, the final assigned index number, the court's treatment of the constitutional theory, and any sworn record about actual policies, audit logs, access lists, contracts, retention exceptions, and case uses. Also pending is a clean public accounting of which outside agencies currently have access, not just which agencies allegedly had access in prior records or through vendor networks.

Disputed or not yet tested: the county had not yet reviewed the suit when AP asked for comment, according to AP's report. A serious article should keep that attached. The county may argue that ALPRs are common law-enforcement tools on public roads, that courts have often allowed plate-reader use, that the system supports public safety, that policies exist, or that the plaintiffs have overstated risk. BadPD's job is not to pre-write the judge's order. It is to keep the receipts organized so the public can see exactly what the county must answer.

The Prior Receipts Make The New Lawsuit More Serious

This case did not land in a vacuum. In March 2025, the Guardian reported on records showing that ICE had accessed license-plate-reader data in sanctuary jurisdictions, including Westchester County. That reporting said Westchester police worked with Rekor, managed a 480-camera network as of January 2023, and recorded 16.2 million scans in the last week of January 2023 alone. The Guardian also reported that ICE, Customs and Border Protection, DHS, the FBI, DEA, Secret Service, and other agencies appeared in records connected to database access as of February 2022, while noting it was unclear whether the access remained current in 2025.

That older reporting matters because the new lawsuit's complaint does not simply say, "This might happen someday." It alleges a built-out system with outside sharing and specific retention practices. The prior records create a public-interest reason to ask for the current access list, current agreements, current audit records, and current retention exceptions.

The complaint also makes a budget and vendor accountability claim. It alleges WCPD contracts with Rekor for software, spent public funds on Rekor Scout access and related maintenance, and separately paid other ALPR vendors for camera equipment. Those details matter because surveillance often expands through procurement before the public understands the policy consequences. A camera contract can become a constitutional fight years later if the board, public, and courts were never given a full vote on the system's real scale.

The clean public-records request path is obvious. Westchester County should be asked for the Rekor contracts, camera inventory, data-retention schedule, hotlist criteria, user access list, interagency access agreements, audit logs, policy manuals, training materials, and annual use statistics. If the county believes the program is lawful and well-governed, those documents are how it proves that without hand-waving.

Why The 99 Percent Figure Is The Core Accountability Hook

The most important phrase in this story is not "AI" and it is not "ICE." It is "over 99 percent." If the complaint's numbers hold, the system's ordinary output is not targeted investigation. The ordinary output is non-investigative movement data on ordinary drivers.

Public-safety agencies will always point to real cases where license plate readers help find stolen cars, missing people, violent suspects, or vehicles tied to serious crimes. Those cases should be weighed honestly. But a working crime tool does not automatically justify unlimited collection, long retention, outside sharing, and warrantless historical searches of everyone else's movements.

The right policy test is not whether ALPRs can ever be useful. The right policy test is whether the county has built guardrails proportional to the power of the tool. That means limits on retention for non-hit data, clear warrant or supervisor-approval standards for historical searches, strict limits on outside sharing, published access lists, audit logs with consequences for misuse, annual public reports, and a governing statute or ordinance adopted by elected officials after public debate.

If those guardrails do not exist, the public-safety argument becomes too easy. Agencies get to keep every success story while ordinary drivers absorb the privacy cost in silence. The county then gets the investigative upside of a mass database without the democratic burden of defending the trade in public.

That is exactly why this lawsuit deserves coverage even before the first major ruling. It creates a forcing mechanism. Either Westchester can show the authority and rules, or it has to defend why a police department should be able to create one of the country's largest vehicle-surveillance systems without them.

The Sharing Problem: Local Camera, National Search

ALPR stories are easy to misunderstand because the cameras are physically local. A driver sees a camera in Westchester and assumes the data is a local record. The complaint challenges that assumption. It alleges that WCPD gives database access to users from at least 55 outside agencies and participates in a vendor network that may broaden access further.

This is the hidden policy step. Local leaders can debate whether their own officers should use cameras at specific intersections, highways, bridges, or border points. But if the same data can be searched by federal agencies, out-of-state agencies, or network partners, then the local debate is incomplete unless it covers downstream access too.

That matters for immigration enforcement, abortion-related investigations, political protests, labor organizing, religious attendance, and medical visits. The complaint's plaintiffs include people who say their vehicle records intersect with work, political organizing, immigration observation, reproductive-health volunteering, and community activity. The lawsuit is not saying each of those activities was actually targeted by WCPD. It is saying the database architecture makes those patterns searchable.

That distinction matters. The scandal does not require proving that every officer misused the system. A system can be poorly governed even before the worst abuse happens. The question is whether the design creates avoidable risk, whether the public authorized the risk, and whether independent checks exist to catch misuse.

Recent national reporting on ALPR misuse makes that concern less theoretical. Separate reporting this week described officers in multiple states accused of misusing Flock license-plate-reader systems to track romantic partners or others for personal reasons. Flock is not the Westchester vendor named in this complaint, and those incidents should not be pasted onto Westchester as if they are the same case. But they show why audit logs and penalties are not paperwork. They are the difference between a public-safety tool and a stalking tool with a badge attached.

The Neighborhood Placement Claim Needs Receipts, Not Slogans

The complaint alleges that Westchester's cameras are heavily concentrated in non-white-majority neighborhoods near the Westchester-Bronx border, where residents are predominantly Black and Latine. That is an important civil-rights claim, but it should be handled as a records question, not as a slogan.

The way to test it is straightforward: publish the camera map, the installation dates, the stated reason for each placement, the board or department approvals, the crime or traffic-safety data used to justify the location, the demographic analysis, and the records showing who requested each camera. If the placement was driven by objective safety needs, the county can show its work. If the placement reflects old patterns of over-policing dressed up as new technology, the records should show that too.

BadPD should not reduce the case to identity rhetoric. The accountable frame is power, authorization, and proof. If a camera network is concentrated in specific communities, those communities deserve a direct explanation of why the cameras are there, how long their data is kept, who can search it, how many searches relate to serious crime, how many searches relate to immigration enforcement, and how misuse is punished.

That is not anti-police. It is the minimum public accounting owed when police build a countywide surveillance system that records people who are not suspected of anything.

What Westchester Should Have To Produce Next

A useful public response from Westchester should not be a one-paragraph reassurance. It should answer the operational questions raised by the complaint and prior records.

First, the county should publish the legal authority it relies on to operate the system at this scale. If the authority comes from state law, county law, board approval, procurement authority, or police executive power, name it. If the Board of Legislators specifically authorized a mass ALPR network with multiyear retention and outside sharing, show the vote.

Second, the county should publish the current data-retention rule for non-hit records. If more than 99 percent of records are unrelated to a hotlist or investigation when collected, the default retention rule is the privacy policy. The public needs to know why two years is necessary, what shorter periods were considered, and who can approve longer retention.

Third, the county should publish the current direct-access list and all data-sharing agreements. If ICE, HSI, FBI, DEA, New York State Police, NYPD, neighboring counties, or out-of-state agencies can search the data, say so. If they cannot, say that too and explain when access ended.

Fourth, the county should publish search statistics. How many historical searches were run last year? How many required a warrant? How many required supervisor approval? How many related to violent crime, stolen vehicles, missing persons, immigration enforcement, narcotics, traffic enforcement, civil matters, or non-criminal intelligence? How many searches produced a case result?

Fifth, the county should publish misuse records. How many users were suspended, disciplined, referred, retrained, or removed? How often does the county audit searches? Are audits random, complaint-based, or automatic? Does the county notify people when their data was misused?

Those records would turn a heated civil-liberties dispute into a measurable public-safety debate. Without them, residents are asked to trust a system that is powerful precisely because most of its work is invisible.

BadPD Accountability Bottom Line

This deserves a full post because it is not just another privacy lawsuit. It is a local government power story with numbers, a primary complaint, a counsel release, independent AP reporting, and prior public-records reporting that gives the allegations context.

The confirmed record is enough to publish carefully: a June 9 class-action complaint challenges Westchester County's ALPR system; the complaint alleges at least 575 cameras, roughly 1.6 billion records, more than 264 million 2024 reads, more than 99 percent non-hotlist data, at least two years of retention, and broad outside access. AP reported the county had not yet reviewed the lawsuit. The court has not ruled.

The missing facts are the county's answer, the final docket posture, the current access list, the contracts, the written policies, the audit logs, and the statutory authority. Those are not side questions. They are the story.

BadPD's position should be simple: if public roads are going to become a searchable historical database of ordinary travel, the public gets to see the rules before the system becomes normal. Westchester can defend targeted public-safety use. It should not be allowed to hide mass retention and broad sharing behind the word "technology."

Reader Source-Status Note

This article treats the lawsuit as a source-backed allegation package, not as a court finding. Westchester County had not yet received or reviewed the lawsuit when AP sought comment, according to AP. The key follow-up receipts are the county answer, current access list, contracts, written policies, audit logs, data-sharing agreements, and court rulings.

Source Trail

• Policing Project case page: Umemoto v. Westchester (Accessed June 13, 2026; case filed June 9, 2026) – Primary case page summarizing the alleged 575 ALPRs, 264,303,687 2024 reads, 1.6 billion reads, retention, no-warrant claim, and agency-sharing allegations.
• Policing Project press release: Westchester drivers sue county police department (June 9, 2026) – Counsel release announcing the class action, named representatives, requested court relief, and source framing from Policing Project, Knight, NYCLU, and Freshfields.
• Complaint PDF: Umemoto v. Westchester (NYSCEF received June 9, 2026; accessed June 13, 2026) – Primary pleading. The PDF includes the New York e-filing caution that the document had not yet been reviewed and approved by the County Clerk at printout.
• Associated Press: Motorists sue Westchester County over 1.6 billion license plate scans (June 9, 2026) – Independent account of the lawsuit, county not-yet-reviewed response, claims about data sharing, plaintiff vehicle-read counts, and broader court context.
• The Guardian: ICE accessed car trackers in sanctuary cities (March 11, 2025) – Prior public-records reporting on Westchester ALPR access, Rekor, ICE/DHS-related records, camera counts, and scan volumes. Used as background, not as proof of current access.
• Tom’s Hardware / 404 Media context: alleged Flock ALPR misuse cases (June 12, 2026) – Separate national ALPR-misuse context. Not a Westchester/Rekor allegation; included only to explain why audit logs and penalties matter.