Flock built an AI tool that tracks drivers by movement patterns

Flock Safety has told the public for years that its cameras cannot identify or track individuals. WIRED reconstructed code from the company's own login pages and found an AI tool that does both. The system, called OS Investigate and previously known as Nightshift, draws on a camera network that logs vehicle movement across more than 6,000 communities. It can surface potential witnesses by counting how often a car passes through a neighborhood, pull a driver's frequent travel companions from cameras they pass together, and convert a license plate into a name, home address, and list of relatives by reaching into arrest records, dispatch logs, and commercial identity databases.

The tool ships with 69 prewritten prompts that officers can select, edit, and submit to the AI, backed by 45 tools that reach plate scans, camera metadata, case files, ballistics results, and commercial records containing Social Security numbers and dates of birth. One prewritten prompt asks the system to find witnesses based on vehicles most seen in a chosen neighborhood over the past 14 days. Another instructs it to list everyone arrested more than twice in two years, map where they live, and produce a background-check "workup" on three of them, selected by the software itself. Fourteen of the prompts require no plate, no name, and no description at all: an officer supplies only a location, a time window, and a behavior, such as visiting three retail locations in three days or passing through multiple gas stations overnight.

“I think that most people imagine that license plate readers are checking license plates and they're looking if a vehicle is wanted. And if it's not wanted, the police will move along. But we're increasingly seeing the aggressive deployment of AI by Flock. And in some ways, like China, very little space between what can be done and what is being done.”

— Jay Stanley, senior policy analyst, ACLU Speech, Privacy, and Technology Project

Flock did not dispute the capabilities WIRED described. Spokesperson Paris Lewbel called OS Investigate a separate product from the company's license-plate reader technology and said its workflows may change before a wider release. The tool is arriving while Flock already faces a documented pattern of misuse. A Texas officer searched more than 83,000 cameras nationwide while looking for a woman who had a self-administered abortion, according to logs obtained by 404 Media. A Washington Post analysis found 50 officers charged with or accused of misusing plate-reader systems, 26 of them to track exes, wives, or women they wanted to meet. Illinois regulators found Flock had violated state law by running a paused pilot that gave Customs and Border Protection access to in-state cameras. The Electronic Frontier Foundation reviewed more than 12 million logged searches and found that roughly 20% listed only a vague reason, like "investigation" or "suspect."

• Communities covered by Flock's camera network: more than 6,000
• Prewritten AI prompts shipped with OS Investigate: 69, backed by 45 data-access tools
• Flock's monthly plate scan volume: 20 billion
• Monthly active users of Flock's platform: roughly 140,000
• Flock's most recent valuation: $7.5 billion

Flock's chief executive, Garrett Langley, has compared the AI system to a crime analyst who is, in his words, always awake. At a demonstration for police chiefs in Denver, he described searching every camera in a city for cars tailing an armored truck and pulling up owners' arrest records in under a minute. He has said investigators who try the tool get "addicted." The company Langley started with a phone in a waterproof box now runs a network large enough that ACLU attorney Chad Marlow, who leads the group's "Get The Flock Out" campaign, worries less about the prewritten prompts than the open-ended ones officers can type themselves. He says an officer asking an AI which people it considers likely to be committing crimes risks building "not a legal standard of criminal suspicion, but whatever that AI standard is."

The tension at the center of this story runs in the opposite direction from where crypto infrastructure has been heading. Flock's system exists to correlate identity across cameras, records, and time. Ethereum's own roadmap has moved toward the reverse goal, with Vitalik Buterin's latest comparison of proposed upgrade paths putting privacy ahead of nearly every other priority for the base layer. Demand for that kind of design has shown up in market pricing too, with privacy coins climbing the token rankings earlier this month. A surveillance tool built to unmask drivers by their driving habits and a blockchain roadmap built to make identity harder to pin down are answering the same question about who gets to know where you've been, just from opposite directions.

Nothing here should be taken as financial advice — just information to consider.