Flock AI Lets Police Search Vehicles by Location, Time, Movement

Flock AI Lets Police Search Vehicles by Location, Time, Movement

I was watching a short video the other night: a grid of tiny car images, timestamps, a cursor hovering over a street name. The realization landed slow and sharp—you can now ask a camera network a question in plain English and get answers about who moved where. I felt the room shrink a little.

I’ve followed surveillance tech for years, and you should care because this isn’t just license-plate reading anymore—it’s a search engine for movement. I’ll walk you through what Wired uncovered, what Flock says, and why the change matters for anyone who drives, dates, or lives in a city.

Camera networks are already everywhere on American streets.

Wired reporters and an independent researcher known as Buchodi found code on Flock Safety’s login portal and used it to reconstruct an interface for a tool called OS Investigate. That reconstruction shows police can query Flock’s camera grid with plain-language prompts and get lists of vehicles that match location, time, and patterns of movement.

You don’t need to be a technologist to see how this shifts the game. Flock’s original pitch was a narrow one: read plates, check them against suspect lists. This tool flips the order—start with a place and a time, and let the cameras surface the cars.

Investigators can now type prompts and get back lists of cars seen at specific times.

The mock interface included sixty-nine prewritten prompts officers can edit. One example asked the system to find “witnesses based on vehicles most seen in [neighborhood] during [last 14 days] during [daily timeframe].” The searches filter out buses, semis, commercial vans, and trailers—so the results zero in on privately owned vehicles.

Critically, the tool asks officers to provide a reason for a search but doesn’t appear to force detailed justifications. You can change the prompt on the fly and submit. It operates like a metal detector sweeping a crowded beach.

Can Flock’s cameras identify faces?

On its website, Flock says its automated license plate readers “do not identify people by their faces” or “identify who is inside a vehicle.” The company told Wired that OS Investigate is “a separate product from Flock’s license-plate reader technology, designed to help investigators work across information their agencies already have access to,” and that the interface could change before public release.

How does OS Investigate let police find drivers?

By stitching plate reads to external records. Wired reports the experimental tool can tie camera hits to 911 dispatch logs, arrest records, and commercial databases—meaning a license plate becomes a pointer to a person and, potentially, their social circle. The camera feed itself may not show a face, but the system makes it trivial to connect dots that previously required work and warrants. The records become a spiderweb of leads.

Police misuse and public backlash have followed Flock’s tools.

There’s precedent. The Washington Post reported that at least fifty officers have been charged with or accused of abusing Flock systems, often to stalk women. That pattern of misuse has produced vandalized cameras, angry city councils, and PR headaches for the company.

Flock announced new safeguards and a user‑behavior flagging tool that customers must adopt by year‑end, according to the Wall Street Journal. Whether those changes will stop casual misuse—or deter a determined insider—remains open.

Departments are buying AI tools while staffing gaps get wider.

Police agencies are under pressure: staffing shortages and budget stress have pushed many to try tech fixes. Firms like Flock Safety and Axon Enterprise are pitching AI features as ways to extend a thin force’s reach. Without federal rules, cities and departments are left to set their own limits.

If you work in local government or follow civil‑liberties fights, this should read like an alarm bell. You and your neighbors deserve policies that match this new capability.

I’m not suggesting cameras will replace courts or detectives. I am saying the searchability of movement—turned into a queryable dataset—changes power on the street, in police cars, and in criminal investigations. That’s authority in code.

So what should you want from policy: mandatory logs that record who searches what, a real audit trail, and strict limits on queries tied to personal reasons? Wired, Buchodi, Washington Post, and others have already raised those questions—are you ready to demand answers?