AI Flagged Her Road Trip 'Suspicious' — and the Officer Showed Her the Map

In 60 seconds: The ACLU says American police are increasingly pulling drivers over not for how they drive, but because AI software flags their travel patterns as "suspicious." One road-tripper was shown a heat map of everywhere she'd been — and the cases behind the warning stretch from Texas to California.

Photo: A City of Alexandria police car equipped with mobile ALPR. File photo via Wikimedia Commons.

Here's a sentence you never expect to hear at a traffic stop: "Let me show you everywhere you've been." That's what Grace Knowles says happened to her — and if the ACLU is right, her story isn't a one-off. It's a preview of what driving in America is becoming.

What happened

Knowles, a twenty-something from northern Virginia, told the ACLU she was on a road trip in September 2025 — Oregon, Texas, Louisiana — when, on her way to Florida after leaving New Orleans, she noticed a police car behind her and got pulled over. She says she was driving at a normal speed. The officer's stated reason, she says, was that she'd been in the left lane too long. Then came the questions: where was she coming from, where was she going, could she prove she'd been in New Orleans? She showed him a pamphlet from a whiskey museum she'd visited. That's when, she says, he relaxed and explained himself: he wasn't a regular cop, he was an FBI special agent, and when he scanned her plate, her travel patterns had come up "suspicious."

Honestly, the detail that sticks with me is what happened next. Knowles says the officer invited her into his car and pulled up the software — a heat map of the United States, with circles marking everywhere she'd been, each one opening a photo of her license plate. She'd never shared her story publicly before telling it to the ACLU, whose post about her case went up on October 26, 2026, and is now trending in US searches. Her account is hers, as told to the ACLU — the FBI hasn't publicly confirmed the stop — but the ACLU pairs it with a stack of cases that rhyme with it uncomfortably well.

Latest verified facts

The technology at the center of this is the automated license plate reader, or ALPR: cameras that photograph passing cars, read their plates, and log the time and place of every read in a searchable database. The ACLU's argument is that AI has changed what that database is for. It used to help investigate a vehicle police already suspected. Now, the group says, it generates suspicion from scratch.

The timeline the ACLU lays out is specific. In July 2025, it says, Flock — the company whose cameras now cover thousands of US communities — was touting an AI ability to alert police not just to stolen cars or wanted plates, but to drivers whose travel patterns looked "suspicious," and to map associations between vehicles frequently seen together. A year later, Wired obtained detailed files on Flock's next-generation AI system, already in use by some police, which the ACLU says enables wide-ranging, open-ended queries across plate data plus arrest records, case files, and commercial data-broker databases packed with personal information on nearly all of us.

Then there are the other cases. The Associated Press reported in November 2025 that the Border Patrol is monitoring drivers nationwide inside the US for "suspicious" travel patterns and working with other agencies to stop and search vehicles — a furniture-moving company driver, for example, was stopped by local police acting at the Border Patrol's direction about 100 miles from the Mexican border after driving from South Carolina; a search of his truck found nothing illegal, but police arrested him on suspicion of money laundering and seized thousands of dollars of his own cash. The ACLU notes it's possible Knowles was stopped under this same program. A Texas jury recently sided with a man who sued his local police department after a pretext stop — the claimed lane drift was contradicted by dashcam footage and later admitted bogus by the officer — because a federal fusion center had flagged his Houston-to-Carrizo Springs round trip as suspicious; the Institute for Justice, which represented him, says the same county subjected thousands of other drivers to the same routine. In March 2026, The Markup reported a California grandmother interrogated by Customs and Border Protection officers about how much time she spends at local casinos, apparently based on her driving patterns. And in April 2025, 404 Media reported a man targeted for driving back and forth between Wisconsin and Michigan — Michigan, local police noted, being "a known source State for Marijuana as it is legal there."

Why it matters

The reason this isn't just a privacy-nerd story is the mechanism: the pretext stop. Police have long pulled people over for a claimed minor infraction — drifting out of a lane, a missed turn signal — as cover for a different real reason. The ACLU says officers even have slang for the data-driven version: "whisper stops" or "intel stops." What AI changes is the scale. A human officer can only follow one car. A camera network watches every car, all the time, and an algorithm decides which ones look odd. The ACLU's blunt warning is that these systems will inevitably sweep in far more innocent people than actual wrongdoers — and that the encounters that follow can be far more frightening than Knowles's, which ended with an explanation and no arrest.

There's also a fairness question hiding in the word "suspicious." A road trip through three states looks suspicious to software trained on routine commutes. Driving to a casino, driving to a state where marijuana is legal, driving a moving truck full of someone else's furniture — all of these are lawful, and all of them have now triggered a stop in the cases above. When the definition of suspicious is whatever the algorithm says it is, there's no way for a driver to know what not to do.

Impact on the USA and the world

For American drivers, the practical upshot is that the road itself has become a sensor network. Flock's cameras are now in thousands of communities, and the data they collect is searchable across agencies and state lines. The ACLU warns the same tools could be turned to political or protest associations and travel — a concern that lands differently in 2026 than it would have a decade ago. Courts are starting to push back in places: a federal judge ruled in February 2024 that a lawsuit challenging Norfolk, Virginia's use of 172 Flock cameras could proceed, finding residents had plausibly alleged the system creates a detailed chronicle of a person's everyday movements. And at least one North Carolina town, Atkinson, has banned the cameras outright.

For the rest of the world, the US is the test case other countries watch. Britain invented plate readers in the 1970s and regulates them under data rules; the US has no warrant requirement and, as the ACLU notes, no real restriction on how law enforcement can use ALPR data. Whatever rules America eventually writes — or fails to write — will shape the market for this technology everywhere.

What happens next

Watch three things. First, whether more drivers come forward with Knowles-style stories now that the ACLU has put a name and a face on the pattern — the group's whole point is that most people never get shown the software, so the true number of AI-flagged stops is unknown. Second, whether courts or legislatures move: the Norfolk case, the Texas jury verdict, and town-level bans like Atkinson's are the early pressure points, and the ACLU is campaigning to ban pretext stops and end plate surveillance outright. Third, what Flock and the agencies say next — the company maintains its cameras help solve crimes and that a plate alone isn't personal information, a position its critics answer by pointing out that a network of reads is a diary, not a data point.

Our take: Knowles got the polite version of this future — an officer who showed her the map and let her go. The cases around hers suggest plenty of drivers get a harder version, with searches, seizures, and arrests built on a pattern an algorithm didn't like. That's the part worth watching, because unlike a speeding ticket, you can't drive more carefully to avoid being flagged. The only fix is a rule about when police get to ask the machine who's suspicious — and right now, America doesn't really have one.

Related reading: How to Get a REAL ID in 2026: Documents, Cost and Airport Rules; ESTA vs eTA vs ETA: US, Canada and Australia Entry Permits, Compared; How to Apply for a Green Card in the USA: Steps, Fees and Timelines.

Sources: American Civil Liberties Union, "Police Are Increasingly Pulling People Over Because AI Deems Their Driving Patterns 'Suspicious'" (Oct 26, 2026); Gizmodo, "Police Are Using AI to Pull Over Drivers for 'Suspicious' Behavior"; Tom's Hardware, "Police reportedly pulling motorists over because AI algorithms deem their driving patterns 'suspicious'"; Associated Press (Border Patrol monitoring report, Nov 2025, via ACLU); The Markup (Mar 2026); 404 Media (Apr 2025); Institute for Justice (Schott case). Knowles's account is as told to the ACLU; the FBI has not publicly confirmed the stop. Flock's position is as reported in the sources above.