DHS, CBP Looking For Facial Recognition Tech That Will Allow Them To Scan Every Person In A Vehicle

(Mis)Uses of Technology

from the adding-known-failure-to-known-failures dept

The festival of bad ideas continues. Despite facial recognition having proven over and over again it’s not really the solution to speedy ID verification that far too many people think it is, government agencies (along with retailers, sports arenas, and bored billionaires) seem to believe the real problem is that there just hasn’t been enough failure.

In 2021, facial recognition tech used by the DHS and its components scanned more than 23 million people. While the number of inputs was depressingly impressive, the end result of all this biometric gathering was merely disappointing:

The report indicates that the system caught no imposters traveling through airports last year and fewer than 100 new pedestrian imposters.

The CBP’s numbers were even worse. More than 50 million face scans between 2018 and 2021, with only a total of 292 imposters caught by the always-on surveillance.

Apparently, the problem is that there isn’t enough of the stuff that already proven itself mostly useless. As Caroline Haskins reports for Wired, the CBP is asking tech firms to whip up facial recognition tech capable of scanning every face in every car crossing the border.

CBP says that its facial recognition tool “is currently operating in the air, sea, and land pedestrian environments.” The agency’s goal is to bring it to “the land vehicle environment.” According to a page on CBP’s website updated last week, the agency is currently “testing” how to do so. The RIF says that these tests demonstrate that while this facial recognition tool has “improved,” it isn’t always able to get photos of every vehicle passenger, especially if they’re in the second or third row.

“Human behavior, multiple passenger vehicle rows, and environmental obstacles all present challenges unique to the vehicle environment,” the document says. CBP says it wants a private vendor to provide it with a tool that would “augment the passenger images” and “capture 100% of vehicle passengers.”

Obviously, this is hardly the ideal environment to produce usable facial recognition images. No matter where the cameras are placed, some passengers’ faces will be obscured in some way, whether it’s by shadows, other people/objects in the vehicle, or simply facing the “wrong” direction.

The thing is, the CBP already knows this isn’t a winning strategy. Haskins’ Wired article quotes the EFF’s Dave Maass, who obtained CBP documents detailing a trial run of this tech from 2021 to 2022. The results of this first deployment were far from impressive.

Maas said that what stood out to him was the error rates. Cameras at the Anzalduas border crossing at Mexico’s border with McAllen, Texas captured photos of everyone in the car just 76 percent of the time, and of those people, just 81 percent met the “validation requirements” for matching their face with their identification documents.

That’s an incredibly high failure rate for tech that was only asked to do “one-to-one” matching — that is, matching a person’s facial recognition scan to identification documents presented to border control officers. With these numbers, you’re better off just relying on the humans staffing border entries to make judgment calls on travel documents.

If this new system is expected to do more than this — like perform one-to-many matches — the error rate will be compounded exponentially. Even if the tech continues to improve, there’s actual life and liberty on the line here. Without rigorous guidelines and lots of human backstops, this tech will just make rights violations more efficient.

It’s not just those coming into the country who will be subjected to this, either. The DHS ultimately wants these camera facing both ways to better capture the faces of everyone in a vehicle leaving the country.

CBP exclusively tells WIRED, in response to an inquiry to the agency, that it plans to mirror the current program it’s developing—photographing every person entering the US and match their faces with their travel documents—to the outbound lanes going to Canada and Mexico. The agency currently does not have a system that monitors people leaving the country by vehicle.

“Although we are still working on how we would handle outbound vehicle lanes, we will ultimately expand to this area,” CBP spokesperson Jessica Turner tells WIRED.

So, even if you’re cool with south-facing cameras scanning every foreigner trying to enter the US, you’re going to have to accept the trade-off no one bothered asking your opinion about. You’re going to be subject to just as much always-on surveillance as people you think aren’t worthy of living in the United States. The Constitution-free zone has never been less free, and it’s only going to get worse, no matter who’s in the White House.

Filed Under: biometrics, border security, cbp, dhs, facial recognition, immigration, surveillance