ALPR Backlash: What Agencies Need to Know

The Big Picture Most of what’s driving the backlash is a governance problem, not a technology problem. Over the summer of 2026, ALPR, and one provider in particular, became one of the most criticized technologies in policing. Cameras were cut down in upstate New York, painted in Oakland, and rammed with a truck in Idaho. Fusion centers circulated bulletins telling agencies to increase patrols around camera installations. Dozens of jurisdictions ended or declined contracts. Vandalizing ALPR equipment seems to have gone viral because of the pushback against the technology. However, the concerns underneath the reaction are a different matter, and most of them are specific: who can search the data and for what purpose, how long it is kept, whether an officer has to give a reason to search, what happens when a misread plate triggers a stop, and whether anyone is checking any of it. Several of these concerns rest on documented events rather than speculation. A state audit found federal immigration authorities had gained access to camera data in violation of state law. Disciplinary records show officers using these systems to track people they knew personally. Wrongful felony stops have been built on misread plates. Unresolved litigation is testing whether warrantless queries of a camera network amount to a Fourth Amendment search. None of this originated on social media, though social media has accelerated public awareness of it. Nearly every one of the concerns directly relates to a setting, a policy, or a supervisory practice rather than a technology-only matter or feature. Critics argue the technology itself is the problem and needs regulation through settings, policy, supervision, and audits. Nearly every documented failure does trace to a setting, a policy, or a supervisory practice. The point of disagreement is not whether those controls are needed; it is that agencies do not have to wait for a legislature to put them in place.” Or if that’s a bit strong: “Nearly every one traces to a setting, a policy, or a supervisory practice, which means the fixes are available to any agency willing to make them. Much of the current discourse is aimed at a single provider, and some of that is fair; national lookup and reciprocal sharing defaults were intentional product choices. But ALPR is not a single-provider category. Agencies have used these tools for years, multiple companies sell comparable systems, and by 2020 roughly 65% of municipal departments with 100 or more sworn officers reported using ALPR of some kind (LEMAS, as cited in Shjarback & Sarkos, 2025). Decisions around retention, query justification, hit verification, and internal auditing impact your agency regardless of who the solution provider is. The Research Base is Thin The formal evidence is a different picture, and less settled than either side of the argument tends to suggest. The strongest experimental work comes from two randomized trials, both of which tested readers mounted on patrol vehicles rather than the fixed networks now under debate. Lum and colleagues (2011) found no reduction in auto theft or crime generally when ALPR-equipped patrols were assigned to hot spots. Taylor, Koper, and Woods (2012) found more stolen vehicle recoveries and auto theft arrests, but no drop in theft itself. A related quasi-randomized test found little evidence that displaying LPRs added to the deterrent value of patrol (Koper et al., 2022). All were part of a broader NIJ-funded evaluation of LPR cost-benefit effectiveness (Koper et al., 2018), the most comprehensive assessment of the technology to date. Studies of investigative use are more encouraging and less conclusive. Koper and Lum (2019) found large-scale LPR deployment may improve clearance rates for auto theft and robbery under some conditions. Shjarback and Sarkos (2025) evaluated a fixed-ALPR expansion covering every entrance and exit to Atlantic City: clearance for shooting incidents improved but not significantly, and while violent crime overall did not fall, the expansion was associated with reductions in shootings, motor vehicle thefts, and property crime. Research on real-time crime centers, where ALPR is one input among several, has found clearance improvements for violent crime (Guerette & Przeszlowski, 2023; Arietti, 2023), though those studies measure the center rather than any single technology inside it. The pattern points more toward investigative value than deterrence. But what has been rigorously tested does not look much like what agencies are buying today. The fixed, upgraded ALPR networks now under debate have not been rigorously evaluated. What the Providers Report Both major vendors publish outcome numbers. One reports its technology is involved in solving roughly 10% of reported crime nationwide, a figure drawn from a 2023 survey of its own customers combined with FBI crime data and extrapolated across the customer base (Flock Safety, 2024). Another reports 8.1 billion plates read in a year, contributing to more than 20,000 felony arrests and 50,000 vehicle recoveries (Axon, 2025). Both providers emphasize the scale of adoption and anecdotal evidence rather than proof of any causal impact. The first provider compares adopting and non-adopting agencies without accounting for why agencies adopt; the second counts activity, not outcomes. Either may be a reasonable way to open a conversation but neither substitutes for measuring what happens in your own agency. When the Hit is Wrong If an officer treats an alert as a fact, two things can go wrong. The reader misreads a character, confusing an O for a zero or a 7 for a 2, or gets thrown by dirt, a tinted cover, or a non-standard plate. Or the read is perfect and the hot list is stale, because a recovered vehicle was never cleared or a suspended registration was reinstated. Error rates in the field are not trivial. A randomized trial in Vallejo, California found that 37% of hits from fixed readers and 35% from mobile readers were misreads (as reported by the Brennan Center, 2020). In Aurora, Colorado in 2020, a reader matched a minivan to a stolen motorcycle registered in another state, and officers held a woman and four children face-down in a parking lot; the city later settled for a reported $1.9 million (CNN, 2024). In Morristown, Tennessee in 2024, an O read as a zero led to a felony stop on two grandparents with their three-year-old granddaughter in the car (WATE, 2025). Safeguards against acting on an erroneous hit predate the current controversy. The Bureau of Justice Assistance policy template states that an alert alone may not rise to reasonable suspicion and is not sufficient probable cause to arrest and directs officers to visually verify the plate characters and state of issuance and confirm the entry is still active before acting (BJA, 2017). The Ninth Circuit reached the same conclusion in constitutional terms, holding that an unconfirmed ALPR hit does not by itself establish reasonable suspicion for an investigatory detention (Green v. City & County of San Francisco, 2014). Where felony stops have gone wrong, the failure has usually been that this step was skipped, not that the camera existed. What the Audits and Discipline Records Show Unfortunately, governance failures are documented far more concretely than the crime impacts of the technology. In August 2025, an Illinois Secretary of State audit of 12 local agencies found its ALPR provider had allowed U.S. Customs and Border Protection to access Illinois camera data in violation of a 2023 state law (Illinois Secretary of State, 2025). Mount Prospect, Illinois disclosed that 262 immigration-related searches had been run against its data by outside agencies through national lookup and opted out of the feature (Village of Mount Prospect, 2025). The Legal Picture Agencies Should Be Watching In January 2026, a federal judge upheld Norfolk, Virginia’s 176-camera network against a Fourth Amendment challenge, finding that the system could not capture the whole of a person’s movements. The same opinion warned that ALPR surveillance could become intrusive enough to cross the line as the technology expands (Courthouse News Service, 2026). Discovery showed the cameras logged the plaintiffs’ vehicle 475 and 325 times in roughly four months. However, on June 29, 2026, the Supreme Court decided Chatrie v. United States, holding 6-3 that obtaining a person’s location history through a geofence warrant is a Fourth Amendment search. The Court rejected the argument that a short window of data falls outside Fourth Amendment protection and narrowed the third-party doctrine along the way (Chatrie v. United States, 2026). That reasoning runs against the limited-scope rationale the district court used to uphold Norfolk’s network. For agencies, the practical read is that the constitutional ground on which ALPR networks rest is unsettled. Norfolk survived on a rationale the Supreme Court has since narrowed, and litigation testing the new reasoning against ALPR networks is only a matter of time. The variables courts have signaled they will weigh, retention length and network scope, are both within agency control. Why Public Support Is an Operational Concern None of this should have caught the field off guard. Researchers surveyed 405 members of the public in 2018 about ALPR and related technologies and found that 90% did not know whether the agency serving their community used the technology at all (Merola, Lum, & Murphy, 2019). Their recommendation was direct: weigh public perception seriously when deciding whether and how to adopt the technology, because awareness would eventually catch up. That awareness has now arrived, and quickly. By most measures June 2026 was an unremarkable month for crime in the United States. However, it was not unremarkable regarding public attention to ALPR: searches for ALPR climbed from the middle of the month to more than 1,000% above where they had been a year earlier, according to Google Trends. That marks the end of operating in an environment where most residents do not know the technology exists. Research on public acceptance of policing technology finds that perceived fairness and transparency drive support at least as much as demonstrated effectiveness, and that how a technology is authorized and overseen shapes public perception independently of what it does (Guler, Kula, & Boke, 2025; Schiff et al., 2025). What Agencies Can Do Now Treat an alert as a lead, not a stop. Require visual verification of the plate characters and state of issuance against the read, and confirmation that the hot list entry is still active, before any enforcement action. Put it in policy, train to it, and check compliance in after-action review of high-risk stops. Set a retention period you can justify and write down the justification. Flock reduced its default retention from 30 days to seven in August 2026 (Flock Safety, 2026). That change applies to new customers; existing customers keep whatever retention period they already have unless they change it. Either way, the vendor’s default is not a justification for your retention period. Look at your own query logs. Look at your own query logs, determine how far back your investigations reach, set the window to match, and document how you got there. That record is what you will need in a council meeting, a public records request, or a suppression hearing. Control who can run a query and require a reason for everyone. Some ALPR vendors provide a search- reason field, but it was free text, and the ACLU has documented officers entering placeholder entries that the system accepted (ACLU, 2026). Case codes and automated audit review are being made mandatory during 2026 (Flock Safety, 2026). The same applies to data sharing: reciprocal sharing and national lookup were the mechanisms behind nearly every cross-jurisdictional problem on record, so pull your network audit and read the list of outside agencies with access to your data. Audit your own use and measure your own results. Review high-volume users and repeat queries on the same plate and treat what you find as a supervision matter. At the same time, record ALPR involvement in case files in a way that separates an early investigative lead from late-stage confirmation, and track your alert accuracy rate alongside it. Most agencies cannot currently say whether a plate read generated a case or merely confirmed it, which is why nobody can settle the effectiveness question. The Bottom Line The evidence on ALPR effectiveness is not where it needs to be. There is limited rigorous research (experimental and quasi-experimental designs) to prove that it works as well as vendors claim and not much showing it doesn’t. Most of what has been tested does not match what agencies are deploying now. The governance record is a different matter. It is specific, documented, and mostly about decisions agencies made or failed to make: how long to keep the data, who to share it with, whether a query requires a reason, whether an officer confirms a hit before drawing a weapon, and whether anyone checked. That distinction explains why swapping vendors resolves less than it appears to. Major cities have ended ALPR contracts and competitor providers have been positioning for that business (NPR, 2026). Some of the criticism was about one provider, and a new vendor addresses it. The rest was about retention, sharing, verification, and oversight, and that must be handled at the agency level not with the vendor. Need assistance? If your agency is reviewing or auditing ALPR use or drafting ALPR policy, NPI can help, as a trusted, independent organization with a 55+ year track record and in-house expertise. Contact NPI. References ACLU. (2026, August). Despite ‘new’ updates, Flock’s creepy cameras remain major civil liberties threat. https://www.aclu.org/news/privacy-technology/despite-new-updates-flocks-creepy-cameras-remain-major-civil-liberties-threat Arietti, R. (2023). Do real-time crime centers improve case clearance? An examination of Chicago’s strategic decision support centers. Journal of Criminal Justice, Volume 90, 2024, 102145, ISSN 0047-2352, https://doi.org/10.1016/j.jcrimjus.2023.102145 Axon. (2025, April 22). Axon announces new fixed ALPR camera solutions and next-gen AI advancements. https://www.axon.com/newsroom/press-releases/axon-announces-new-fixed-ALPR-camera-solutions-and-next-gen-AI-advancements-to-expand-real-time-public-safety-ecosystem Brennan Center for Justice. (2020). Automatic license plate readers: Legal status and policy recommendations for law enforcement use. https://www.brennancenter.org/our-work/research-reports/automatic-license-plate-readers-legal-status-and-policy-recommendations Bureau of Justice Assistance. (2017). License plate reader policy development template for use in intelligence and investigative activities. https://bja.ojp.gov/sites/g/files/xyckuh186/files/media/document/LPR_Policy_Template_FINAL_2-7-170.pdf Chatrie v. United States, 609 U.S. ___ (2026). https://supreme.justia.com/cases/federal/us/609/25-112/ CNN. (2024, February 5). Aurora, Colorado, will pay $1.9 million settlement after officers drew weapons on Black family in a stolen vehicle mix-up. https://www.cnn.com/2024/02/05/us/colorado-aurora-settlement-stolen-vehicle-mixup/index.html Courthouse News Service. (2026, January 27). Judge holds Norfolk’s license plate reader use constitutional. https://www.courthousenews.com/judge-holds-norfolks-license-plate-reader-use-constitutional/ Flock Safety. (2024). How many crimes do automated license plate readers (ALPRs) solve, anyway? https://www.flocksafety.com/customers/how-many-crimes-do-automated-license-plate-readers-alprs-solve-anyway Flock Safety. (2025). 1,000 missing persons reunited through Flock Safety. https://www.flocksafety.com/blog/1000-missing-persons-reunited Flock Safety. (2026, August 13). Flock updates privacy, accountability, security, and transparency safeguards. https://www.flocksafety.com/blog/flock-guardrails-address-lpr-privacy-concerns-and-police-transparency Green v. City & County of San Francisco, 751 F.3d 1039 (9th Cir. 2014). https://caselaw.findlaw.com/court/us-9th-circuit/1666208.html Guerette, R. T., & Przeszlowski, K. (2023). Does the rapid deployment of information to police improve crime solvability? A quasi-experimental impact evaluation of real-time crime center (RTCC) technologies on violent crime incident outcomes. Justice Quarterly, 40(7), 950–974. https://www.tandfonline.com/doi/full/10.1080/07418825.2023.2264362 Guler, A., Kula, S., & Boke, K. (2025). Examining public support for AI in policing: The role of perceived procedural justice. Police Practice and Research, 26(6), 673-695. https://doi.org/10.1080/15614263.2025.2516535 Illinois Secretary of State. (2025, August 25). Giannoulias’ audit finds license plate reader company in violation of state law. https://www.ilsos.gov/news/2025/august-25-2025-giannoulias-audit-finds-license-plate-reader-company-in-violation-of-state-law.html Koper, C. S., & Lum, C. (2019). The impacts of large-scale license plate reader deployment on criminal investigations. Police Quarterly, 22(3), 305-329. https://journals.sagepub.com/doi/abs/10.1177/1098611119828039 Koper, C. S., Lum, C., Willis, J. J., Happeny, S., Johnson, W. D., Nichols, J., Stoltz, M., Vovak, H., Wu, X., & Nagin, D. S. (2018). Evaluating the Crime Control and Cost-Benefit Effectiveness of License Plate Recognition (LPR) Technology in Patrol and Investigations. Final Report to the National Institute of Justice. Fairfax, VA: Center for Evidence-Based Crime Policy, George Mason University. Koper, C. S., Lum, C., Wu, X., Johnson, W., & Stoltz, M. (2022). Do license plate readers enhance the initial and residual deterrent effects of police patrol? A quasi-randomized test. Journal of Experimental Criminology, 18, 725–746. https://doi.org/10.1007/s11292-021-09473-y KRQE. (2026, August). Albuquerque police: License plate readers have helped solve 13 murder cases in 2026. https://www.krqe.com/news/albuquerque-metro/albuquerque-police-license-plate-readers-have-helped-solve-13-murder-cases-in-2026/ Lum, C., Hibdon, J., Cave, B., Koper, C. S., & Merola, L. (2011). License plate reader (LPR) police patrols in crime hot spots: An experimental evaluation in two adjacent jurisdictions. Journal of Experimental Criminology, 7(4), 321-345. https://www.ojp.gov/library/publications/license-plate-reader-lpr-police-patrols-crime-hot-spots-experimental Merola, L. M., Lum, C., & Murphy, R. P. (2019). The impact of license plate recognition technology (LPR) on trust in law enforcement: A survey-experiment. Journal of Experimental Criminology, 15(1), 55-66. https://doi.org/10.1007/s11292-018-9332-8 NPR. (2026, August 22). As Flock battles public scrutiny, other police surveillance companies see an opening. https://www.npr.org/2026/08/22/nx-s1-5931446/flock-fallout-competitors-see-opportunity Redwood City Police Department. (2026). Automated license plate reader (ALPR) program. https://www.redwoodcity.org/departments/city-manager/automated-license-plate-reader-alpr-program Schiff, K. J., Schiff, D. S., Adams, I. T., McCrain, J., & Mourtgos, S. M. (2025). Institutional factors driving citizen perceptions of AI in government: Evidence from a survey experiment on policing. Public Administration Review, 85(2), 451-467. https://doi.org/10.1111/puar.13754 Shjarback, J. A., & Sarkos, J. A. (2025). An evaluation of a major expansion in automated license plate reader (ALPR) technology. Justice Evaluation Journal, 8(2), 225-242. https://doi.org/10.1080/24751979.2025.2473363 Taylor, B., Koper, C., & Woods, D. (2012). Combating vehicle theft in Arizona: A randomized experiment with license plate recognition technology. Criminal Justice Review, 37(1), 24-50. Village of Mount Prospect. (2025, June). Police department response to Flock license plate reader investigation. https://www.mountprospect.org/Home/Components/News/News/10311/1042 WATE. (2025, June 11). License plate mix-up leads to couple being handcuffed in Morristown. https://www.wate.com/news/top-stories/couple-handcuffed-in-morristown-due-to-license-plate-mix-up/ Never miss an issue of InFocus Share