Secret Financial Surveillance: How Border Patrol’s Predictive Teams Are Feeding Police Tips

Imagine a highway patrol officer pulling you over because a computer somewhere flagged your recent purchases as “suspicious.” That’s not a dystopian novel—it’s happening, according to a recent 404 Media investigation. The agency behind the move? The U.S. Border Patrol, specifically a little‑known unit called the Predictive Intelligence Targeting Team, or PITT.

PITT doesn’t actually chase down drivers. Instead, agents scan data streams that are supposed to be restricted to law‑enforcement use—credit‑card transactions, cryptocurrency moves, even receipts from legal dispensaries. When the algorithm spots a pattern that looks like the financial fingerprints of drug trafficking, the team sends a nudge to the nearest local police department.

That nudge can become a traffic stop. In Montana, a man named Kyle William Olson was pulled over after a PITT analyst in nearby Spokane flagged his bank activity. The officers told Olson they stopped him because his license plate was partially obscured—a perfectly routine reason, the kind you’d expect on any road. Yet the real catalyst was a spreadsheet of his recent purchases, including legal cannabis edibles he’d bought from a farm in California where he worked.

Olson was charged with a DUI and an alleged intent to distribute narcotics, even though the cannabis he carried was lawful in the state where it originated. The case illustrates a worrying shift: law‑enforcement agencies are using financial data as a pretext for stops, bypassing the traditional requirement of “probable cause.”

Experts aren’t shy about calling this practice a legal sleight‑of‑hand. Jake Laperruque, deputy director of the Security and Surveillance Project at the Center for Democracy and Technology, warned that “genuine probable cause cannot be synthetically generated.” He likened the process to “parallel construction,” a technique where the original investigative trail is hidden, replaced by a manufactured justification.

What’s more opaque is the technology itself. The Border Patrol’s public statement speaks in vague terms about “intelligence‑informed analysis” used to “support national security operations.” It refuses to disclose the exact algorithms, data sources, or how long they retain the financial snapshots they collect—citing “operational security reasons.”

This secrecy makes it nearly impossible for the public, journalists, or even courts to evaluate whether the surveillance is proportionate, accurate, or legal. As Laperruque put it, without meaningful oversight, trust evaporates.

Critics argue that expanding predictive policing into the realm of personal finance is a massive escalation. Earlier iterations of predictive policing relied on crime‑statistics maps, which, while imperfect, at least dealt with public data. Now, a citizen’s spending habits—something many consider private—can trigger a police stop, even if the purchases are entirely lawful.

The case of Olson is just one flashpoint. The investigation suggests that dozens of similar stops have occurred across the country, each originating from a quiet data analysis in a PITT office. Whether the intention is to catch drug traffickers or simply to widen the net of surveillance, the result is the same: ordinary Americans find themselves under the lens of a system that they never consented to.

Calls for transparency are growing. Civil‑rights groups want clear rules about what financial data can be accessed, how it’s stored, and—crucially—who gets to decide that a pattern is “suspicious.” Until those safeguards are in place, the line between legitimate security work and invasive monitoring will remain dangerously blurry.