This AI Watches Millions Of Cars Daily And Tells Cops If You’re Driving Like A Criminal

Artificial intelligence is helping American cops look for “suspicious” patterns of movement, digging through license plate databases with billions of records. A drug trafficking case in New York has uncloaked — and challenged — one of the biggest rollouts of the controversial technology to date.

By Thomas Brewster, Forbes Staff

In March of 2022, David Zayas was driving down the Hutchinson River Parkway in Scarsdale. His car, a gray Chevrolet, was entirely unremarkable, as was its speed. But to the Westchester County Police Department, the car was cause for concern and Zayas a possible criminal; its powerful new AI tool had identified the vehicle’s behavior as suspicious.

Searching through a database of 1.6 billion license plate records collected over the last two years from locations across New York State, the AI determined that Zayas’ car was on a journey typical of a drug trafficker. According to a Department of Justice prosecutor filing, it made nine trips from Massachusetts to different parts of New York between October 2020 and August 2021 following routes known to be used by narcotics pushers and for conspicuously short stays. So on March 10 last year, Westchester PD pulled him over and searched his car, finding 112 grams of crack cocaine, a semiautomatic pistol and $34,000 in cash inside, according to court documents. A year later, Zayas pleaded guilty to a drug trafficking charge.

“With no judicial oversight this type of system operates at the caprice of every officer with access to it.”

Ben Gold, lawyer

The previously unreported case is a window into the evolution of AI-powered policing, and a harbinger of the constitutional issues that will inevitably accompany it. Typically, Automatic License Plate Recognition (ALPR) technology is used to search for plates linked to specific crimes. But in this case it was used to examine the driving patterns of anyone passing one of Westchester County’s 480 cameras over a two-year period. Zayas’ lawyer Ben Gold contested the AI-gathered evidence against his client, decrying it as “dragnet surveillance.”

And he had the data to back it up. A FOIA he filed with the Westchester police revealed that the ALPR system was scanning over 16 million license plates a week, across 480 ALPR cameras. Of those systems, 434 were stationary, attached to poles and signs, while the remaining 46 were mobile, attached to police vehicles. The AI was not just looking at license plates either. It had also been taking notes on vehicles' make, model and color — useful when a plate number for a suspect vehicle isn’t visible or is unknown.

To Gold, the system’s analysis of every car caught by a camera amounted to an “unprecedented search.” “This is the specter of modern surveillance that the Fourth Amendment must guard against,” he wrote, in his motion to suppress the evidence. “This is the systematic development and deployment of a vast surveillance network that invades society’s reasonable expectation of privacy.

“With no judicial oversight this type of system operates at the caprice of every officer with access to it.”

Gold declined to comment further on the case. Westchester County Police Department did not respond to requests for comment.

Reckoning with Rekor

Westchester PD’s license plate surveillance system was built by Rekor, a $125 million market cap AI company trading on the NASDAQ. Local reporting and public government data reviewed by Forbes show Rekor has sold its ALPR tech to at least 23 police departments and local governments across America, from Lauderhill, Florida to San Diego, California. That’s not including more than 40 police departments across New York state who can avail themselves of Westchester County PD’s system, which runs out of its Real-Time Crime Center.

“You've seen the systems totally metastasize to the point that the capabilities of a local police department would really shock most people.”

Brett Max Kaufman, senior staff attorney at the ACLU

Rekor’s big sell is that its software doesn’t require new cameras; it can be installed in already deployed ones, whether owned by the government, a business or a consumer. It also runs the Rekor Public Safety Network, an opt-in project that has been aggregating vehicle location data from customers for the last three years, since it launched with information from 30 states that, at the time, were reading 150 million plates per month. That kind of centralized database with cross-state data sharing, has troubled civil rights activists, especially in light of recent revelations that Sacramento County Sheriff’s Office was sharing license plate reader data with states that have banned abortion.

“The scale of this kind of surveillance is just incredibly massive,” Brett Max Kaufman, senior staff attorney at the ACLU, told Forbes. Pointing to both Rekor and Flock, a rival that runs a similar pan-American surveillance network of license plate readers, he described warrantless monitoring of citizens en masse like this as “quite horrifying.”

Rekor declined an interview and did not respond to requests for comment. But Matt Hill, who sold his company OpenALPR to Rekor in 2019 and was its chief scientist before departing in September 2022, told Forbes the network was likely growing. “I’m fairly sure there would be more cameras in more states now… It would be fairly large,” he said, noting that non-government customers have also been providing video. Rekor’s private customers include parking lots, casinos and fast-food chain restaurants.

With so many agencies now collecting license plate records, and the dawn of more advanced, AI-powered surveillance, privacy advocates are raising the alarm about a technology expanding with little in the way of legal protections for the average American. “You've seen the systems totally metastasize to the point that the capabilities of a local police department would really shock most people,” added Kaufman. “This is just the beginning of the applications of this technology.”

‘Unavoidable surveillance’

The ALPR market is growing thanks to a glut of Rekor rivals, including Flock, Motorola, Genetec, Jenoptik and many others who have contracts across federal and state governments. They’re each trying to grab a slice of a market estimated to be worth at least $2.5 billion.

But it’s not easy. Reporting first quarter results for this year, Rekor saw $6.2 million in revenue with a $12.6 million net loss. It reported a similar loss in the same quarter last year. Its stock is currently trading at about $2.75 down from an April 2021 high of $23.45 per share

In pursuit of that elusive profit, the market is looking beyond law enforcement to retail and fast food. Corporate giants have toyed with the idea of tying license plates to customer identities. McDonalds and White Castle have already begun using ALPR to tailor drive-through experiences, detecting returning customers and using past orders to guide them through the ordering process or offer individualized promotion offers. The latter restaurant chain uses Rekor tech to do that via a partnership with Mastercard.

With the sheer breadth of expansion, it’s becoming increasingly difficult to avoid the watchful eyes of government and corporate surveillance — or even know where they are. As Gold found in trying to get data from the Westchester government, authorities are not legally obliged to provide information on the whereabouts of cameras.

“Given the vast nature of the ALPR network and the need to travel public highways to engage in modern life,” Gold wrote in his motion to suppress, “avoiding ALPR surveillance is both unfeasible, if impossible.”