Federal Agents Are Quietly Mining Your Bank Records to Justify Police Stops

Imagine driving down a highway, minding your own business, when a patrol car flashes its lights for a supposedly "obstructed license plate." You roll down the window, and before you can even say "what's the problem," officers pull you out of the car, search you, and hand you a citation. It sounds like a routine traffic stop, right? What you probably didn’t expect is that the reason you’re being detained could have started miles earlier—in a spreadsheet full of credit‑card transactions.

That’s the unsettling reality uncovered by 404 Media, which traced a covert operation inside the U.S. Border Patrol. The agency has created specialized units called Predictive Intelligence Targeting Teams, or PITT, that mine what they call “law‑enforcement‑sensitive systems” for financial patterns they associate with drug trafficking, smuggling, or other illicit activity. The teams don’t do the stops themselves; instead, they whisper their findings to local police departments, who then make the traffic stop seem completely ordinary.

One concrete example involves a man named Kyle William Olson, who was traveling across Montana. According to the report, a PITT analyst in nearby Spokane, Washington—Border Patrol Agent Matthew Phelps—noticed that Olson’s credit‑card purchases matched “financial activity patterns commonly associated with illicit narcotics activity.” The analyst passed that intel along to Montana sheriffs, who later stopped Olson on the pretext of a faulty license plate. During the stop, they discovered cannabis edibles in his vehicle, seized them, and slapped him with DUI and drug‑distribution charges. Olson says the weed came from a licensed farm in California where he works, not from any underground operation.

Critics are already sounding the alarm. Jake Laperruque, deputy director of the Center for Democracy and Technology’s Security and Surveillance Project, told 404 that “genuine probable cause cannot be synthetically generated.” He warned that the Border Patrol is essentially using “parallel construction” to hide the true motive behind these stops, making it impossible for anyone to challenge the legality of the arrest.

The agency’s official line is, of all things, vague. In a brief statement, Border Patrol said it employs “intelligence‑informed analysis and planning” to protect national security and public safety, but declined to disclose the exact tools, data sources, or algorithms it uses, citing “operational security reasons.” In other words, they won’t tell you how they’re watching your bank statements, just that they’re doing it.

It raises a host of privacy questions. If a federal unit can scan your grocery purchases or your Netflix subscription and then tip off a state trooper, where does the line between legitimate threat detection and unconstitutional surveillance get drawn? The practice is a dramatic escalation of predictive policing—a method already under fire for amplifying bias and relying on opaque statistical models.

For now, the mechanics remain murky. No public documents detail the specific databases or AI models PITT relies on, and there’s no clear legal framework governing the sharing of this kind of financial intelligence with local law‑enforcement partners. What’s certain, though, is that the trend signals a new frontier in the surveillance state—one where your wallet might become the next checkpoint.