Plate Readers: Results, Misuse, Courts and Pushback

Police chiefs who buy plate readers tell stories: the stolen pickup found in an hour, the kidnapping suspect caught on the interstate, the hit-and-run driver identified from a partial plate and a dented fender. Those stories are often true. Privacy advocates tell different stories: the officer who tracked an ex-girlfriend hundreds of times, the out-of-state search that a state law was written to stop. Those are often true too. A fair look tests both kinds against records.

What the research says about results

The research base is thinner than the sales pitch and thinner than the warnings. Start with the best-designed work on each end.

The first is a 2011 randomized experiment by Cynthia Lum, Christopher Koper and colleagues at George Mason University, published in the Journal of Experimental Criminology. They took 30 auto-crime hot spots in two adjacent jurisdictions and randomly assigned 15 to get patrols with plate-reader cars. The patrols did not produce a measurable deterrent effect on auto theft or theft from autos, either during the experiment or in a 30-day follow-up. The authors cautioned that the experiment used small numbers of patrols and suggested better ways to use the equipment.

The second is a 2026 working paper by Scott Mourtgos and Ian Adams that compared 216 agencies that deployed Flock cameras from 2017 to 2023 with more than 3,000 that did not. It found vehicle theft fell about 11 percent after deployment and motor vehicle theft arrest rates rose about 15.9 percent. Recovered cars were found about seven hours sooner at the median. The Institute for Justice, which is suing Norfolk, Virginia over its cameras, published an explainer listing the limits: theft was already falling in adopting agencies, adoption was not random, and the 11 percent drop was statistically significant only when results were weighted toward agencies with high prior theft. A working paper has also not yet been through peer review.

Investigations are a separate question from deterrence. Agencies often keep anecdotal lists of solved cases. Ask for the list, then ask for the case numbers and check a sample. A department that can document 40 cases in which plate data was a meaningful lead has given you something real. A department that cannot name one has given you something too.

Documented misuse

The misuse record has grown fast. The Institute for Justice keeps a running list of officers reported to have used plate readers to track romantic partners, exes or people they were fixated on. As of October 2026 it lists 81 incidents and says the real number is likely higher. In Lufkin, Texas, an officer was indicted on 100 felony counts after allegedly searching for an ex-girlfriend and people she knew more than 10,000 times. A Savannah, Georgia officer was fired after allegedly making 400 searches of an ex-girlfriend and another person. A Milwaukee officer allegedly tracked a woman he was dating and her former partner nearly 180 times in two months. In August 2026 the Washington Post counted at least 50 officers accused of or charged with plate reader misuse, 46 of them involving Flock. Flock told the Post misuse is a small fraction of its roughly 140,000 active users.

Idaho has its own case. Flock logs obtained by KMVT showed more than 700 searches for Jerome County Sheriff George Oppedyk's wife's vehicle between early July and mid-September 2025, all labeled "test." A letter from the Idaho Attorney General's Office put it at about 152 queries of her plate in one month and said the sheriff's stated purpose was checking the system's reliability before a contract decision. A lieutenant who manages the system flagged the searches. The chief deputy reported them to the Attorney General and to the Twin Falls police chief, whose department shares cameras with the county. Twin Falls Police Chief Matthew Hicks said his department could not find "any legitimate reason" for one person to run one plate that many times as a test. The Attorney General declined to prosecute, saying there was not sufficient evidence to prove a crime beyond a reasonable doubt and that policy violations were outside his office's authority. Twin Falls cut off the sheriff's office's access to its cameras for about 48 hours, until the sheriff's own search access was removed and a new policy adopted. Oppedyk retired April 18, 2026, about halfway through his term. He did not respond to KMVT's calls and emails.

Misuse is not only personal. The Electronic Frontier Foundation analyzed more than 12 million Flock searches logged by more than 3,900 agencies from December 2024 to October 2025. It found more than 50 federal, state and local agencies running hundreds of searches tied to protests, including the No Kings demonstrations. The Spokane County Sheriff's Office, for one, entered "no kings" as the reason for three searches on June 15, 2025, across 95 camera networks. Illinois passed a 2023 law barring the sharing of plate data with agencies investigating out-of-state abortions or immigration status. In 2025 the Illinois Secretary of State found U.S. Customs and Border Protection had accessed Illinois plate data anyway, and a suburban Illinois department had shared data with a Texas sheriff searching for a woman who had a self-managed abortion. On August 25, 2025, Flock paused pilot programs with federal agencies.

Notice how most misuse comes to light: a complaint, a victim checking public audit tools, a supervisor who happened to look, or a state official pulling the logs. The tool records every search. What fails is the habit of reading those records. The California State Auditor's February 2020 review of four agencies (Fresno, Los Angeles, Marin County and Sacramento County) found none had audited staff searches, and three shared images widely without checking whether recipients had a right to them.

The courts: the tide is turning on mass collection

For years, plate reader challenges failed on a simple idea: a plate on a public road is in plain view. That idea is cracking, because courts now look at what a network of cameras adds up to, not at one photo.

The foundation is Carpenter v. United States (2018), where the Supreme Court held that getting weeks of a person's cell phone location records is a search that needs a warrant, even though each data point was held by a company. In 2021 the full Fourth Circuit, in Leaders of a Beautiful Struggle v. Baltimore Police Department, held that Baltimore's aerial surveillance program, which photographed most of the city from planes during daylight, violated the Fourth Amendment because it let police reconstruct people's movements after the fact. And on June 29, 2026, in Chatrie v. United States, the Supreme Court held that pulling a person's Google Location History through a geofence warrant is a search.

Plate readers are now getting the same treatment in some courts:

The other side still has the larger stack of rulings. On January 27, 2026, U.S. District Judge Mark Davis ruled for Norfolk in the Institute for Justice's challenge to its network of close to 200 Flock cameras, writing that courts have "almost uniformly concluded" that photographing plates on public roads is not a search. In September 2026 a federal judge in California held that querying the Flock database is not a search either. Judge Hill herself acknowledged that nearly every court to look at the question had gone the other way. Her ruling binds no other court, and Flock said it expects it to be overturned on appeal.

So the honest summary: courts have found that collecting and searching everyone's movements without a warrant can violate the Fourth Amendment, in the Baltimore plane case, in Carpenter and Chatrie, and now in at least one federal plate reader ruling and one jury verdict. No appeals court has yet held that a plate reader network itself is unconstitutional. Watch the Fourth Circuit, where the Norfolk appeal is headed, and the Tenth Circuit if the Oklahoma ruling is appealed.

Cities pulling out

Local governments are canceling or refusing contracts at a pace that was hard to picture two years ago. Counts differ because trackers define a "cancellation" differently, so name the tracker every time:

• Secure Justice, a Bay Area anti-surveillance group, counted 214 cities and towns that had dropped Flock since 2021, 90 of them in August 2026 alone (reported Aug. 31, 2026).
• Newsweek, analyzing DeFlock data, counted 149 cities and counties that canceled or rejected contracts or shut off cameras from January through August 2026.
• DeFlock.org listed 153 cities that had rejected Flock cameras as of early October 2026.

Idaho is part of it. Post Falls announced on July 24, 2026 that it would not operate Flock cameras. Salmon passed a resolution saying it would not acquire plate readers "at this time." The Jerome City Council removed Flock funding from its 2027 budget over privacy concerns. Clearwater County's sheriff said in July he would not renew, citing a lack of "sufficient operational value." Wilder amended its contract after a tort claim, then declined to adopt a citizen-proposed ban, which sends the question to Wilder voters in November 2027. Twin Falls' mayor has signaled the city could modify or cancel its contract. Most large Idaho cities still use the cameras: Twin Falls has 40, Pocatello 36 and Idaho Falls 20, according to the Idaho Capital Sun.

DeFlock and the citizen response

DeFlock is an open-source map where anyone can view and report the locations of plate reader cameras. It turned camera locations, which agencies often did not publish, into public information. EastIdahoNews used it to count 46 cameras in the Pocatello area and 19 in the Idaho Falls area.

DeFlock Idaho is a separate, decentralized group of Idaho residents opposed to the cameras. It says it does not coordinate with the national project and does not condone vandalism. Organizer Brian McKellar filed a $100,000 tort claim against Idaho Falls in July 2026 for invasion of privacy, saying he wants the cameras down, not money. Organizer Tyler Cain filed tort claims against Caldwell and Wilder, arguing that the contracts let Flock use residents' data for non-law enforcement purposes, which Idaho Code 49-1432 forbids. Wilder amended its contract, and Idaho Falls said it would amend its contracts so Flock cannot use the data to train its AI. The group has also used public records requests to obtain Ada County Sheriff's Office audit logs.

Candidates are taking positions too. Sen. Tammy Nichols (R-Middleton) has said the Legislature must put "privacy, accountability and Fourth Amendment protections first," and has named shorter retention, warrant requirements or a ban as options. Several legislators and candidates have signed privacy pledges opposing mass surveillance, among them Rep. Clint Hostetler (R, District 24A) and the Republican nominees in District 26, Bryan Hyde for Senate and Lyle Johnstone for House Seat B. For anyone following this issue, a signed pledge is a record: save a copy, note the date, and compare it with each signer's votes once the 2027 session starts on January 11.

In the news: video and local coverage

Recent Idaho coverage of the plate reader debate, listed for reference and not as endorsements:

• "Deflock Idaho reacts to Wilder not banning ALPRs", KTVB (YouTube video).
• "Who's Watching You? Deflock Idaho's Fight Against AI Surveillance", The Idaho Sheriff channel (YouTube video).
• Video on Idaho cities dropping Flock Safety cameras and replacing them with less conspicuous brands, Idaho Freedom Foundation (Facebook video).
• DeFlock Idaho, the Idaho residents' group's own site, with local updates, an explainer on how the cameras and Flock's shared network work, and a link to the camera map.

State laws

The National Conference of State Legislatures counted at least 28 states with laws that expressly address plate readers or their data as of September 2026. Here is a sample.

What Idaho's law says

Idaho added Idaho Code 49-1432 in 2025 (Senate Bill 1180, 2025 session laws chapter 316), effective July 1, 2025. It lets state and local agencies use plate readers for law enforcement, traffic flow analysis or both. Law enforcement use is limited to investigating felonies, misdemeanors and traffic accidents and searching for missing or endangered persons. The readers may not be used to enforce traffic infractions unless an accident is involved. Only authorized personnel may get the data, and it "shall not be used for any non-law enforcement purpose." Agencies must train users, log every access and query with user identity, date, time and purpose, and audit those logs every six months.

What the statute leaves out matters just as much. It sets no retention period. It says nothing express about sharing with out-of-state or federal agencies. Whether its authorized-personnel language limits outside searches has not been tested in court, as far as published sources show. For any Idaho resident, the semiannual audit requirement is a gift: an agency using plate readers since mid-2025 should have audit results to produce.