Flock OS Investigate AI Can Identify and Track Drivers

Flock OS Investigate AI Can Identify and Track Drivers

Flock OS Investigate can identify drivers and track vehicles by their patterns of movement alone. That's a stark departure from the company's long-standing public promise that its technology "cannot recognize, identify, or track individuals." But here's the reality: the vehicle surveillance giant has built an AI tool for police, drawing on a network of cameras logging movements in more than 6,000 communities. It can pick out potential witnesses by how often their cars pass through a neighborhood. It can surface a driver's "associates" from the cameras they pass together. And it can turn license plates into names, home addresses, and relatives by reaching police case files, 911 dispatch logs, and commercial identity records. That's the whole picture. It's stark.

The software, originally called Nightshift and recently renamed OS Investigate, ships with 69 prewritten prompts that officers can select, review, edit, or submit to the AI. Officers can also input prompts of their own. The suggested prompts sit in a cache of more than 450 files found on Flock’s own website, served by its login pages to anyone who loaded them. The code describes 45 tools at the AI’s disposal, giving it access to plate scans, camera metadata, arrest records, case files, dispatch logs, ballistics results, and commercial databases containing Social Security numbers, dates of birth, phone numbers, email addresses, relatives, and associates.

Flock says it's testing the product with a small group of law enforcement partners. It's still in development, they stress. Capabilities may not reflect what it eventually sells. But the announcement lands as the company faces bipartisan political pressure, a growing record of officers caught misusing its platform, and a wave of vandalism that has left cameras sawed off and lenses painted over in cities across the country. That's a lot of heat.

The Inversion of a Simple Transaction

Flock's business model rests on a simple transaction. A camera photographs a passing car, converts the plate to text, and checks it against a list of vehicles police are looking for, while cars that aren't on the list are recorded and otherwise ignored. OS Investigate's preloaded prompts describe an inversion of that arrangement. An officer no longer needs a plate, a name, or a crime to begin. Instead, they supply a place, a stretch of time, and a pattern of behavior, and the system is designed to hand back the people who fit. That's a profound shift.

All 69 prompts are offered to officers as a menu of canned searches. It's a simple trick. Selecting one pastes it into a chat box, where the officer can review or edit it before clicking again to submit, and that's where the real power lies. One prompt preloaded into the system reads: “Find me witnesses based on vehicles most seen in [neighborhood] during [last 14 days] during [daily timeframe] *(will not include whitelisted vehicles).” The output is a list of plates, which other tools in the product then convert into names and home addresses. So the whole thing turns a vague hunch into a doorstep in seconds. That's the kicker.

Another prompt tells the system to list everyone arrested more than twice in two years for “any offense,” exempting only narcotics arrests. Then it maps where those people live. It retrieves calls for service at their homes, and it does a workup on the top three individuals. That’s the key step. A “workup” is a one-command background check that starts with a name and date of birth and returns vehicles, prior listings as a suspect, relatives, phone numbers, and online accounts, so the whole process can run in seconds. But it doesn’t stop there. The system pulls all that data together, and you can’t just glance at it; you have to understand what it means. So the prompt’s design is blunt, and its reach is wide.

Pattern Hunting Without a Target

19 of the prompts describe hunting for patterns rather than looking up a record. 14 call for no plate, no name, and no description. The officer simply supplies a location, a window of time, and a behavior. One group asks for vehicles that visited three or more retail locations in a city in three days, or multiple banks in a week, or multiple gas stations between midnight and 5 am. None of the five names a target or an incident. The tool carries a filter that strips out buses, semi trucks, work vans, and trailers, so a delivery driver visiting five shops would come off the list while an ordinary driver making the same rounds remains.

Most people imagine license-plate readers just check if a vehicle is wanted. That’s it. If it’s not wanted, the police move along, says Jay Stanley, a senior policy analyst with the ACLU’s Speech, Privacy, and Technology Project. But we’re increasingly seeing aggressive AI deployment by Flock, and in some ways, like China, there’s very little space between what can be done and what is actually being done. It’s a thin line now.

Associates and the Two-Minute Window

One example takes a single plate and asks the system to plot the journeys of that car and its “top three associates.” The code describes explicitly how the ranking works: it counts how often other plates appear at the same cameras within a two-minute window of the target by default, and reports back any car that shows up next to it three or more times within a “confidence threshold” defaulting to 0.75. An officer supplies one vehicle and the software finds up to 20.

“I don’t know how else to say this, but this sounds completely insane,” says Noel Pichardo. He was briefed on Flock’s technology during a pilot program while serving as a Pawtucket, Rhode Island, police officer, and he reviewed the prompts. It’s that simple. “I don’t know how anyone can argue against the idea that Flock literally tracks people.” Pichardo spoke out against the technology before Rhode Island state lawmakers in 2024, weeks after serving a 30-day suspension for criticizing Flock and department leadership in a local newspaper. The state police chiefs’ association publicly rebuked him days later. So he resigned as the department moved to fire him. And that’s the whole story.

“We’ll use whatever tool is available to solve crime. It isn’t controversial.”

That quote comes from a detective at a California law enforcement agency, granted anonymity because he wasn't authorized to speak publicly. The detective said the witness-finding prompts made him slightly uncomfortable. But he added, “As law enforcement, we need to use every tool available, in spite of what the ‘defund’ people say.” It's a blunt stance. And it's one that sits uneasily with his own admission of discomfort, a tension he doesn't fully resolve before moving on to justify the practice as a necessity in the face of public criticism.

What the Code Actually Shows

WIRED analyzed the software by examining code Flock served from its own login portals. Loading those pages returned a user interface for screens that would ordinarily appear only after a police user signed in, letting reporters reconstruct portions of the authenticated applications, their prompts, search fields, tools, and controls, without actual access. The files appeared to contain everything needed to reconstruct the interface from scratch.

Before an officer runs a search, the software can ask for a justification. But the code shows that by default an officer has to type a reason, and the form places no real requirements on what that reason says or how long it must be. Just three characters. If a department also requires a case number, the form only checks that it contains at least that many. So it's a low bar. WIRED couldn't determine whether Flock applies additional checks on its servers after the information is submitted, and that uncertainty leaves the system's true rigor unclear.

Flock didn't answer our questions. And they didn't dispute the reported capabilities either. Instead, spokesperson Paris Lewbel framed OS Investigate as something entirely separate from the company's license-plate-reader technology, casting it as a tool meant to help investigators navigate and synthesize information their agencies already possess, which is a subtle but important distinction from the surveillance hardware that made Flock famous. He added that the tool's workflows and features could shift dramatically before any wider rollout. That's a big caveat.

The Constitutional Questions

Andrew Guthrie Ferguson, a law professor at George Washington University, says a system like this one was inevitable. To use all the data collected, you need to build in the ability to query it, he explains, because raw information without a search tool is just noise. He adds that police officers won't create their own agentic tools, so you must preprogram the system with natural language prompts and automated workflows to make it easier for them. But the choice of prompts is what strikes him. There are a lot of innocent activities that could be flagged, even if it also might be quite useful for investigators, he notes. That's the tension.

Chad Marlow runs the ACLU's "Get The Flock Out" campaign. He's an attorney, and he says the real danger lies in the open-ended nature of the prompts. The difference between a police officer saying, "Do you see any criminal patterns" or "Find me criminal patterns" is going to produce two different things, he explains, because the first asks for observation while the second demands an active search. It's a subtle shift, but it matters. And he's worried there are no limitations on the prompts at all. That enables the police to almost lead the AI in what it does, he says, essentially steering the tool toward whatever conclusion they might already suspect. So the system can't just answer; it can be directed.

Scale and Scrutiny

Flock has been building the AI system under scrutiny. Police have already abused the cameras it runs. A Texas officer searched more than 83,000 cameras across the country last year, hunting for a woman who had a self-administered abortion, according to search logs. But the recent Washington Post analysis found 50 officers charged with or accused of misusing plate reader systems, and 26 of them used the tech to track wives, girlfriends, exes, and women they wanted to meet. Illinois regulators found last August that Flock had broken state law. It ran a now-paused pilot program, handing Customs and Border Protection access to cameras in the state. That's the pattern. It's hard to ignore.

Flock leaves most of the rules to its customers. That's the whole point. Agency leaders decide who gets credentials, whether to share their camera data across jurisdictions, and what limits to place on searches, so the power sits with the people using the system, not with some corporate headquarters dictating policy from afar. The company counts roughly 140,000 monthly active users and logs 20 billion plate scans a month. That's a staggering number. And Andreessen Horowitz, Founders Fund, and other venture firms have poured money into the business, which was valued at $7.5 billion in its most recent funding round, even as the product's ethical boundaries remain deliberately fuzzy.

• 69 prewritten prompts ship with OS Investigate
• 45 tools described in the code
• 450+ files found on Flock’s own website
• 20 billion plate scans logged monthly

Garrett Langley, Flock’s chief executive, once likened the tool to a crime analyst who is, in his words, “always awake, always watching out for you.” At a demonstration for police chiefs in Denver last year, Langley described the ability to search every camera in a city for cars tailing an armored truck, pulling up the registered owners of those vehicles and checking their arrest records, in under a minute. Investigators who try it, he said, get “addicted.”

Stanley says the system closely resembles what the US Supreme Court took up in June, when it held that people keep a privacy interest in records of where they have been, and that the Fourth Amendment applies when police demand those records from a company. Here, police departments buy that capability outright, opening space for fishing expeditions on a scale he says has no precedent. “You’ve got a literal geofence. But the AI can also bring it way beyond that kind of a static boundary, and who is inside that boundary,” he says. “Who was near this person for more than 10 minutes in the past month? Who was in these six locations more than three times? It can become much more fine-grained, sophisticated, flexible.”

Frequently Asked Questions

What is the primary difference between Flock OS Investigate and the company's previous public promise about its technology?

Flock OS Investigate can identify drivers and track vehicles by their patterns of movement alone, which is a stark departure from the company's long-standing public promise that its technology 'cannot recognize, identify, or track individuals.' The article highlights this as a significant shift from their earlier stance.

How does one of the preloaded prompts in OS Investigate help officers find witnesses?

One preloaded prompt says: 'Find me witnesses based on vehicles most seen in [neighborhood] during [last 14 days] during [daily timeframe] *(will not include whitelisted vehicles).' The output is a list of plates, which other tools in the product then convert into names and home addresses, turning a vague hunch into a doorstep in seconds.

What does the 'workup' command do in the OS Investigate system?

A 'workup' is a one-command background check that starts with a name and date of birth and returns vehicles, prior listings as a suspect, relatives, phone numbers, and online accounts. This allows the whole process to run in seconds, as described in the article.

According to the article, what are the concerns raised by Chad Marlow regarding the open-ended prompts?

Chad Marlow says the real danger lies in the open-ended nature of the prompts, as the difference between asking 'Do you see any criminal patterns' versus 'Find me criminal patterns' produces different results. He worries there are no limitations on the prompts, enabling police to lead the AI toward whatever conclusion they might already suspect.

What does the article reveal about the default justification and case number requirements before an officer runs a search?

The article states that before an officer runs a search, the software can ask for a justification, but by default an officer has to type a reason with no real requirements on what it says or how long it must be, just three characters. If a department requires a case number, the form only checks that it contains at least that many characters, making it a low bar.