The infrastructure of American policing is shifting from reactive enforcement to predictive surveillance, and the results are unsettling for everyday drivers. A new report from the ACLU details how automated license plate reader (ALPR) networks, powered by AI algorithms from vendors like Flock, are generating 'suspicious' travel patterns that trigger pretextual stops. The real-world impact is already visible: drivers are being pulled over for minor, often fabricated infractions because an algorithm flagged their route as anomalous.

The Grace Knowles Incident

In September 2025, Grace Knowles, a young traveler from northern Virginia, was stopped by an officer who claimed she was driving suspiciously. Knowles told the ACLU she was doing a simple road trip through Oregon, Texas, and Louisiana. The officer, who identified himself as FBI, admitted he had scanned her plate and seen 'suspicious travel patterns.' He even invited her to view the software interface, which displayed a heat map of her movements across the US. This incident highlights the opacity of these systems; the 'suspicion' was not based on driving behavior, but on a data point generated by a commercial surveillance platform.

Flock’s AI and the Expansion of Scope

The technology driving these stops is rapidly evolving. By July 2025, Flock was touting its ability to use AI to alert law enforcement to drivers with unusual travel habits. A year later, detailed files obtained by Wired revealed that Flock’s next-generation system can perform open-ended AI queries on license plate data, cross-referencing it with arrest records and commercial data broker information. This isn't just about catching stolen cars; it’s about creating suspicion where none existed. The system tracks 'associations' between vehicles, effectively building a graph database of who is seen with whom on the road, and then feeding those connections back into the stop decision logic.

Nationwide 'Whisper Stops'

This isn't an isolated glitch. The Associated Press reported in November 2025 that Border Patrol is monitoring drivers nationwide for 'suspicious' patterns, leading to what police internally call 'whisper stops' or 'intel stops.' In one case, a furniture mover was stopped 100 miles from the border after driving from South Carolina; his cash was seized despite no illegal items being found. In Texas, a jury sided with a man stopped because he drove from Houston to Carrizo Springs, a route deemed suspicious by a federal fusion center. These cases demonstrate how AI-driven data pipelines are enabling law enforcement to bypass traditional probable cause, relying instead on algorithmic output that often mistakes normal travel for criminal behavior.

Key Takeaways

  • AI-driven ALPR systems are moving beyond identifying stolen vehicles to predicting 'suspicious' behavior based on travel patterns.
  • Pretextual stops are becoming the primary tool for acting on these AI predictions, often citing minor or invented traffic violations.
  • Vendors like Flock are integrating commercial data broker information with license plate reads, creating a comprehensive surveillance profile for nearly every driver.
  • Legal challenges are emerging, with juries and ACLU lawsuits contesting the constitutionality of stops based solely on algorithmic 'suspicion.'

The Bottom Line

We built the pipes for real-time location tracking, but we failed to write the access controls. Now, an algorithm decides if your road trip looks like a crime, and you have no way to audit the code that stopped you.