Flock’s AI Surveillance Tool Tracks Drivers by Movement Alone—No License Plate Required
Flock Safety's OS Investigate system can identify drivers using movement patterns and behavior alone, contradicting the company's claims about what its technology can do.
Movement-Based Tracking Without Identifying Information
Flock Safety, the surveillance company operating license-plate readers across more than 6,000 U.S. communities, has deployed an AI tool capable of identifying and following individuals based solely on their movement patterns. According to Wired’s investigation, which obtained the system’s underlying code from files on Flock’s website, the tool operates far more invasively than the company has publicly acknowledged.
Called OS Investigate (formerly Nightshift), the system equips law enforcement with 69 pre-built AI prompts that query vast databases including vehicle registrations, arrest records, and commercial data repositories containing Social Security numbers and family relations. Most striking: fourteen of these prompts require no license plate, name, or personal description whatsoever. An officer supplies only a location, a timeframe, and a behavioral pattern—vehicles visiting three retail stores within three days or multiple banks in a week—and the system returns matching drivers’ identities, automatically filtering out commercial vehicles to keep ordinary civilians in view.
The tool’s “associate-finding” function maps social networks through proximity alone. It identifies vehicles appearing at the same camera locations within two-minute windows, flags those appearing three or more times above a 0.75 confidence threshold, and returns up to twenty names. Flock has long claimed its technology “cannot recognize, identify, or track individuals,” yet this functionality does precisely that through movement data.
Minimal Oversight, Expansive Reach
Forty-one of the 69 prompts require no special permission to execute, raising questions about departmental oversight. The query library reveals extraordinary capabilities: one identifies individuals arrested more than twice in two years (excluding drug charges) and builds dossiers on the top three; another flags the vehicles most frequently seen in a neighborhood over two weeks as “witnesses”; a third transforms a name and birth date into relatives, phone numbers, and online accounts drawn from police records and commercial brokers.
Flock confirmed OS Investigate remains in development, currently testing with select law enforcement partners.
The Gap Between Promise and Reality
Flock’s public assurance that the technology is “not for watching people” now clashes sharply with documented capabilities. Noel Pichardo, a former Pawtucket, Rhode Island police officer, reviewed the prompts and stated: “I don’t know how anyone can argue against the idea that Flock literally tracks people.”
Jay Stanley, a senior policy analyst at the American Civil Liberties Union, warned that the gap between Flock’s claims and actual capabilities leaves “very little space… like China,” equating U.S. surveillance infrastructure to authoritarian regimes.
The findings illustrate how commercial surveillance systems evolve from passive logging into active tracking—a transformation highlighting why crypto’s emphasis on financial privacy and self-custody continues gaining traction against centralized systems capturing behavioral and financial data.
Source: Wired, via Decrypt. Not financial advice.