Dayton Bagged 72 Cameras It Can't Remove

Introduction

Dayton, Ohio suspended its Flock camera program on May 1 after an internal review found the city's own data had been searched roughly 7,100 times for immigration-related reasons by outside agencies, in direct violation of city policy. The cameras went dark. Then city workers started climbing up to all 72 of them and covering each one with a black plastic trash bag. Deputy City Manager Joe Parlette told the city commission that the police department would "work with Public Works to put bags over the cameras," because nobody was sure the city could actually take them down under the contract it signed.

The Bags Are the Tell

The bags were a stopgap. The cameras were already switched off, so covering them did nothing technical. What the bags did was answer a question the city couldn't answer any other way: how do you visibly prove to residents that surveillance has stopped when the hardware belongs to a company that may not let you remove it? Dayton appropriated $30,000 for an independent audit of its Flock data logs, and Commissioner Darius Beckham said plainly, "I join my colleagues in requesting the Flock cameras be taken down." He requested. He couldn't order, because the contract didn't give the city that power.

Here's the part that turns one city's mess into a pattern. When Evanston, Illinois terminated its Flock contract in August 2025, Flock removed the bag-covered cameras, then allegedly reinstalled them without Evanston's permission. The city found fresh data logged after the supposed removal in Flock's own public-facing portal, sent a cease-and-desist letter, and re-covered the cameras a second time. Flock didn't pull that second batch until early 2026.

Flock built a surveillance network designed for expansion, not for democratic exit. When a community votes to leave, it discovers the hardware stays in the ground, the contract doesn't include a real termination mechanism, and the company can keep the network node alive while everyone argues. Tom Bowman of the Center for Democracy and Technology told Fortune that cities are "caught in a contractual trap where the cities have signed multimillion dollar contracts with Flock without realizing that they lacked legal or technical mechanisms to force Flock to actually shut down the cameras upon cancellation." Those disputes, he said, "have left cameras active during months of bureaucratic and legal gridlock." I keep coming back to that phrase. The cameras stay on while the lawyers fight, and Flock is the one who controls the off switch.

What Cities Signed Without Reading

Dayton's problem didn't surface in a council chamber. A new police division commander noticed an "unexpected level of data sharing" in October 2025, and the Dayton Daily News investigation that followed found that law enforcement agencies across the country run thousands of searches of local police data every single day. In January alone, hundreds of those searches listed "immigration" as the reason. The commander who was supposed to implement the safeguards that would have blocked this, per Police Chief Kamran Afzal, no longer works there.

This is where the consent gap lives. Flock's defense, repeated in nearly every story, is that "local agencies own and control their data, data sharing is off by default, and data is deleted after 30 days by default." That "off by default" line does a lot of work. Sharing stays off only if someone at the city actively keeps it off and configures the system correctly. Dayton's commander didn't, and the searches flowed for months before anyone caught it. The architecture put the burden of saying no on a single municipal employee, and when that employee failed, the data moved exactly the way the network was built to let it move.

Dayton wasn't a cheap mistake to make. The city got approval in 2024 to spend $825,750 on 35 new cameras and took a $90,000 grant for 27 more earlier in 2026, right up until it pulled the plug. And Ohio gives cities almost nothing to push back with. As of April 2026 there were no state laws governing automated license plate readers, which ACLU of Ohio's Gary Daniels described to WOSU as "quite literally the Wild West." A WOSU records investigation confirmed 305 Flock cameras across just 15 Columbus-area departments, costing nearly $2 million.

Wisconsin's Second Lock

Now go up to Wisconsin, where the exit trap has a second lock that has nothing to do with Flock's contract language. Communities there have been leaving fast. WPR reported on June 3 that Verona, Sturgeon Bay, Oshkosh, Fitchburg, and others have ended their Flock contracts since late 2025. Dane County supervisors voted 32-1 in April to strip $80,000 in camera funding from the budget, and Sheriff Kalvin Barrett confirmed on May 30 that he'll comply. Oshkosh discontinued its contract after concluding Flock had "misrepresented" what its data could do.

Even how Dane County got its cameras tells you something about how thin the oversight was. The sheriff installed the Flock system there without board approval in the first place, funded by a $68,750 grant from Axon Enterprise, a competing surveillance vendor and Flock business partner. The elected body that voted 32-1 to defund the cameras was voting on something it never approved buying.

Milwaukee wanted to do something too. After a Milwaukee police officer was criminally charged in February for allegedly running a romantic partner's plate through Flock 179 times over two months, a coalition of alders moved to pass a community-control ordinance regulating how the department uses surveillance tech. Their own city attorneys told them they couldn't. The reason is 2023 Wisconsin Act 12, a shared-revenue law passed to bail out Milwaukee's pension crisis. Buried in that fiscal rescue was a provision that stripped the Milwaukee Fire and Police Commission of its general policymaking authority and blocked the Common Council from setting surveillance policy, handing that control to the police chief instead. Alder Alex Brower put it bluntly: "We cannot propose that law here. It was extremely frustrating to find that out."

Nobody passed Act 12 to protect Flock. It was a pension fix. But the fine print moved surveillance oversight out of reach of the elected body residents would normally lobby, and the result is that Milwaukee's 31 cameras and its $182,900 contract running through January 2027 sit beyond the council's regulatory reach even after a documented stalking case. So the trap has a second layer. Where Dayton's contract blocks the exit, Milwaukee's state law blocks the city from even writing the rule in the first place. The federal level offered no backstop either: the House Transportation and Infrastructure Committee defeated an amendment to the BUILD America 250 Act that would have barred recipients of federal highway funding from using ALPRs for anything but tolling.

A third version played out in Humboldt County, California on June 3, where supervisors held a four-hour hearing. More than 40 residents spoke, and all but two asked the county to end its Flock contract. The board voted 4-1 to form a committee to study alternatives and directed Public Works to draft a policy on camera placement, but the motions that would have actually pulled funding from the sheriff's $21,000-a-year contract and banned the readers from county property both died for lack of a second. Sheriff William Honsal argued the board has no authority over his "investigative tools." Supervisor Steve Madrone countered that the board controls the budget. Four hours of near-unanimous public opposition produced a study committee and not much else, which is its own kind of exit trap: the cameras stay up while the question gets referred.

Who Benefits

Every camera that stays up is the point, even a bagged one. Flock's value isn't really the individual camera in Dayton or Evanston. It's network density. A camera in any city adds a searchable node to a national database, and the more nodes there are, the more valuable the platform becomes to every other paying customer. So a bagged, switched-off camera that Flock still physically controls is still an asset, still a contract with murky termination rights, still leverage. Keeping it in the ground during months of "legal gridlock" is the business model working exactly as designed.

The receipts back this up. Flock raised $275 million in March 2025 at a $7.5 billion valuation led by Andreessen Horowitz, and its annual recurring revenue passed $300 million, up 70% year over year. By April 2026, Sacra Research pegged its implied valuation at roughly $8.4 billion. That kind of growth runs on subscription renewals and network scale, which means every city that cancels is a direct hit to the asset, and every camera that lingers after cancellation softens the blow. Andreessen Horowitz, which has now backed Flock through three rounds and also funded the drone startup Flock bought, has a financial stake in the cameras staying exactly where they are.

Federal agencies benefit from the same opt-out design that burned Dayton. Immigration enforcement never needed a contract with Flock. It could route requests through cooperative local departments, which is precisely what the 7,100 searches of Dayton's data suggest happened. The "off by default" setting that depended on a city employee actively maintaining it gave federal agencies access at arm's length, with a layer of deniability built in. That's covered ground from the prior pieces in this series, so I'll leave it there, but it's the same architecture producing the same outcome.

When You Can't Hold the Ground, Go Up

Here's Flock's answer to 57 cities (as of late April 2026) canceling contracts and workers bagging cameras: drones. In October 2024 the company acquired a 17-month-old drone startup called Aerodome for more than $300 million, which Aerodome's founder called one of the largest tech acquisitions in public safety history. By September 2025 Flock had launched a private-sector drone product whose coverage spec is 38 square miles per unit, with 45 minutes of flight time and deployment in seconds. Toledo, which is keeping all 107 of its Flock cameras fully operational despite Dayton's suspension, asked its council on June 2 for additional drone funding.

The money is already moving. Stockton, California's city council voted on April 1 to approve a $3,150,000 amendment adding a Drone as First Responder (DFR) platform, dispatched via 911 rather than private alarm, extending the city's Flock contract all the way to 2031. Oakland County, Michigan commissioners approved a drone pilot on April 8, before opening the floor to public comment. The drone push predates the current backlash, so it isn't purely a reaction to it. But the logic lines up too neatly to ignore. Ground cameras face votes, bags, and cease-and-desist letters. Airspace dodges all three. A drone isn't something a city worker can climb up and cover with a trash bag.

The other side of this deserves a hearing, because reading what the record actually says cuts both ways. The revolt is real but it's a fraction of Flock's footprint, which the company puts at more than 5,000 law enforcement agencies. DeFlock.org's tally of 57 cities canceling (as of late April 2026) sits against a network the company says serves thousands more. The public-safety case isn't fabricated, either. Toledo's police chief and Humboldt's sheriff both point to solved cases, and a federal judge in Virginia ruled in January that Norfolk's use of Flock didn't violate the Fourth Amendment. A city that locked its settings down properly may have avoided Dayton's problem entirely. None of that changes the structural point: the cities that did want to leave are the ones who found out there was no clean way to do it.

And aerial coverage locks customers into longer, larger contracts at exactly the moment the cheaper ground product is generating revolt. Stockton signed up for six years and $3.15 million while other cities were busy bagging their cameras. The strategy is already paying off.

The Bottom Line

The Dayton trash bags read at first like a small, almost funny municipal failure. Look closer and they're the clearest evidence yet that the exit was never part of the design. A city can vote to cancel, appropriate money for an audit, and pass a resolution requesting removal, and still end up with a public works crew covering company-owned hardware in plastic because legal removal isn't clear. In Wisconsin, a pension law quietly took the decision away from the people residents would normally petition. At the federal level, the one amendment that might have created leverage got voted down in committee.

More cities will try to leave. The open question is what they find when they do. If the lesson from Dayton and Evanston is that cancellation triggers months of gridlock with the cameras still controlled by the vendor, the smarter move is to never sign in the first place. So the real fight moves past the cameras already in the ground to a simpler question: will any city sign the next drone contract knowing the only exit it's been shown so far is a trash bag?