Leaked documents from surveillance company Perceptics show its pitch to New York's MTA on how it could automatically track and profile vehicles and drivers
Just months before millions of its internal documents were stolen and dumped on the internet, the Tennessee-based surveillance company Perceptics … Tweets: @samfbiddle , @theintercept , @rogerdhodge , @theintercept , @kvnweb , @joejerome , @angelsdiaz_ , @michaelwprice , and @torproject Tweets: Sam Biddle / @samfbiddle : Perceptics, the recently hacked border surveillance firm, recently pitched New York's MTA on a plan that would involve automatically tracking and profiling cars in the heart of Manhattan—and their drivers https://theintercept.com/... pic.twitter.com/l0nUwQF5e2 @theintercept : Perceptics' system is an entirely new sort of surveillance apparatus that tracks deeply personal information like “customer travel patterns and travel consistency,” the number of passengers in the car, or “likely trip purpose.” https://theintercept.com/... Roger D. Hodge / @rogerdhodge : “Leaked Perceptics emails show that the company shipped camera hardware to the MTA's Bridges and Tunnels division for a live demonstration.” https://interc.pt/2XD6QCb by @samfbiddle @theintercept : “To solve a relatively minor problem and maybe save a few bucks, you are creating a system that's going to surveil New Yorkers and present huge risks to their privacy.” https://interc.pt/2NJvCfA Kevin Webb / @kvnweb : I'm 100% certain the “smart city” dystopia will be built under gov't contract. So many are arguing publicly that this kind of tech is necessary (it isn't), and privately that “privacy is dead” (only true if we decide to kill it) https://theintercept.com/... Joseph Jerome / @joejerome : “You've heard of #facialrecognition; this is car recognition.” https://theintercept.com/... pic.twitter.com/SZMjdOyjG1 @angelsdiaz_ : I'm quoted in this @samfbiddle piece about Perceptics' pitch to profile cars, their drivers/passengers, and travel patterns for @MTA. Not only is this an overbroad solution for NYC's congestion pricing, it's an attractive data set for ICE. https://theintercept.com/... Michael Price / @michaelwprice : To implement congestion pricing in #NYC, @MTA being pitched “an entirely new sort of surveillance apparatus that tracks deeply personal information” like travel patterns, consistency, number of passengers, “likely trip purpose.” Not ok. https://theintercept.com/... @torproject : A border surveillance firm, which has been hacked, wants to profile drivers and passengers in New York and track “customer travel patterns and travel consistency” and “likely trip purpose.” https://theintercept.com/...
Context & Ripple Effects
Perceptics, a Tennessee border-surveillance vendor, had already built its business on automated license-plate recognition when it pitched New York's MTA on extending that capability into the heart of Manhattan — not just reading plates but profiling vehicles and their drivers. The pitch only became public because the company was itself breached, with millions of internal documents stolen and dumped online.
The story lands mid-arc in a run of New York surveillance reporting: months earlier, an [[a:940644|NYT project showed Amazon's facial recognition service matching people from city cameras for about $60]], and within a year Clearview AI would claim hundreds of law enforcement agencies using its scraped-image search. The MTA is a recurring customer in this arc — by 2023 it would be running AI-powered surveillance against fare evaders in seven subway stations, suggesting the agency's appetite outlasted the controversy around this pitch.
First-order effects
- The leak hands MTA officials and New Yorkers a look at a proposal that was never publicly debated, and named privacy experts including Joseph Jerome and Michael Price immediately flagged the driver-profiling element as the red line.
Second-order effects
- Every vendor selling plate-recognition or biometric systems to transit agencies now has to answer for Perceptics' pitch — procurement conversations shift from capability to what data is retained, shared, and sold.
Third-order effects
- If the pattern holds — cameras plus cloud matching, as with Clearview AI and the NYT experiment — the structural endpoint is municipal infrastructure doubling as a profiling layer over everyone who drives or rides, with regulation arriving after deployment rather than before.
The trend: Public transit agencies are becoming quiet customers of an expanding surveillance-vendor stack, moving from license plates toward full behavioral profiling faster than oversight bodies can respond.