Cleveland's Flock fight is no longer just about a $250,000 camera contract. It is a test of whether local governments can govern surveillance tools that were built to scale faster than public oversight.
Cleveland Mayor Justin Bibb has moved the city's Flock Safety renewal to City Council after pressure from council members, residents and privacy advocates. That matters because the contract, which covers about 100 license plate readers and expires at the end of June, was close to being handled through the city's Board of Control, the same route used for Cleveland's recent ShotSpotter renewal.
The immediate issue is not only process. It is access. As Signal Cleveland reported this week, audit logs released by the city showed almost 250 searches from outside Cleveland that included terms such as immigration, ICE, Border Patrol, Homeland Security or customs. The logs do not prove that federal immigration officers directly accessed Cleveland's cameras. They do show how difficult it is for a city to know, and explain, what happens once its local data is connected to a national police technology network.
Public Safety Director Wayne Drummond has said Cleveland does not allow immigration searches and that filters have been in place since November. Activists with Flock No CLE are not satisfied. Their argument is simple: if public records and audit logs are the only way residents learn how the system is being used, then oversight is already running behind the technology.
Why Flock's network changes the local debate
License plate readers are often sold as local public safety tools. A camera catches a stolen car, police get an alert, and a case moves faster. That is the clean version. The harder version is that Flock's system can also connect agencies across cities and states, creating a searchable network that a local council may not fully understand when it approves a contract.
Flock says communities own and control their data, sharing with federal agencies is disabled by default, and the company does not work with ICE or any other Department of Homeland Security sub-agency. Those distinctions matter. They also do not answer the full governance question. If a local agency chooses broad sharing, or if an outside officer searches through a participating network, the practical result can still be access that residents never imagined when the cameras went up.
Cleveland is not dealing with this in isolation. Shaker Heights recently changed its Flock policy after records showed hundreds of immigration-related lookups by outside agencies. Dayton went further, suspending its automated license plate reader program after officials said they found more than 7,000 immigration-related search requests and policy violations. These are not abstract privacy worries. They are operational failures showing up in audit logs.
The problem with blaming configuration
The phrase that should make every city attorney pause is configuration error. It sounds technical, harmless and fixable. But when a surveillance system is used by dozens or hundreds of agencies, configuration is policy. A default setting can do more than a council resolution. A sharing toggle can matter more than a public hearing.
That is why Cleveland's council fight is bigger than one vendor. If officials say immigration searches are banned, residents should be able to see how that ban is enforced, who reviews violations, how quickly access is suspended and whether outside agencies are removed when they break the rules. A policy that cannot be audited is closer to a promise than a control.
Flock has added safeguards, including keyword blocks, audit tools and clearer controls around federal sharing. The company also says every search is logged and tied to a user account. Those are useful improvements. But they still rely on local agencies to configure the system correctly, review usage and act when something goes wrong. Technology can create logs. It cannot make a city read them.
What surveillance startups should learn
Flock is now a major public safety technology company, not an experimental startup selling cameras to a few suburbs. The company raised $275 million in 2025 at a $7.5 billion valuation and said it had crossed $300 million in annual recurring revenue. It also says its technology is used by more than 6,000 communities nationwide.
That scale is exactly why the Cleveland dispute matters for founders selling AI and data infrastructure to government. Growth creates governance debt. A tool can spread from city to city faster than procurement rules, public records systems and council oversight can adapt. When that happens, every unexplained search becomes a trust problem, even if the vendor can point to a policy page saying the customer remains in control.
For civic AI companies, the lesson is not to avoid government work. It is to build public accountability into the product from the start. Audit logs should be easy for agencies to publish. Sharing settings should be understandable to nontechnical officials. Contract renewals should come with usage reports, violation histories and clear evidence of public value. If a city cannot explain who searched its data and why, the product has outgrown the oversight around it.
Flock No CLE wants Cleveland to end the contract. Bibb's administration wants to keep the system. Council now has to decide whether the benefits are strong enough and whether the controls are real enough. Whatever it chooses, the market should pay attention. AI surveillance will keep moving into city budgets, but the next fight will not be only about cameras. It will be about whether the public can see the systems watching it.
Also read: MIT builds Fractal OS to find M1 security flaws • SoftBank rides Nvidia's AI surge as investors double down on concentrated bets • Binance's Android app privacy fight is back in the spotlight
