AI-powered surveillance costs lives at the southern border

SPEAKER_1 (0:00)

This Marketplace podcast is presented by Tomorrow's Cure. If our episodes on the intersection of AI and healthcare left you wanting to learn more about how technology is transforming medicine, Tomorrow's Cure is for you. It's the chart-topping 2025 AMBE Award finalist podcast from Mayo Clinic. Back for a brand new season with new host award-winning journalist Lindsay Sievert, Tomorrow's Cure explores the innovations changing the healthcare landscape. Featuring conversations with leading physicians, researchers and medical experts, the new season examines everything from AI-powered diagnostics and cutting-edge cancer therapies to surgical technologies improving patient care today.

Not sure where to start? Listen to the season premiere featuring MD Anderson radiation physicist Dr. Paige Taylor and Mayo Clinic radiation oncologist Dr. Adam Holtzman. They discuss why carbon ion therapy is generating excitement in the medical community and what it could mean for the future of precision cancer treatment. So go ahead, follow Tomorrow's Cure on Apple Podcasts, Spotify, or wherever you get your podcasts.

Megan McCarty-Carino (1:05)

The border is watched by hundreds of surveillance cameras. So how do migrants keep dying in plain view? From American Public Media, this is Marketplace Tech. I'm Megan McCarty Carino.

Over the last 25 years, the US government has spent billions to build a virtual wall of video surveillance at the southern border. The network can signal Border Patrol to respond when it detects suspicious activity, and officials have touted the benefit of providing aid to migrants in distress. But a recent investigation by MIT Technology Review with the Times of San Diego found more than a thousand migrants have died over the last decade within range of one of these cameras. Here with more is Eileen Guo, who co-reported the series Dying on Camera.

Eileen Guo (2:02)

There were the General Dynamics Remote Video Surveillance Systems, which had cameras and thermal imaging on poles.

Then there were Elbit's Towers specifically built in Arizona, which added radar to the cameras to detect motion. Then in 2018, the agency began to roll out Andoril's AI-powered autonomous surveillance towers. What these were meant to do was make it easier to automatically detect and classify the targets, whether it was people or vehicles that were moving through. But again, in all of these generations, we found that the technology was just failing to prevent the fatalities.

Megan McCarty-Carino (2:43)

And your team actually did topographic analysis to determine sort of the sight lines for cameras and whether the location of these deaths should have been in the tower's view. Tell me more about that.

Eileen Guo (2:57)

Yeah, that's right. We worked with a really incredible researcher in the UC system that looked at whether or not, you know, by the tower height, by the topography of where the remains were found, whether or not someone would have been able to see the tower.

And we did find that in some areas, like 10 to 15 percent of the deaths that occurred were visible to towers. And so that brings up the question of, well, why are these towers being placed where they are if there's so much that they cannot see because of the topography?

Megan McCarty-Carino (3:32)

So what are the implications of these failures? What could be going wrong here?

Eileen Guo (3:36)

I think one of the really big questions for us is whether the towers were able to detect that someone was there or not. Or the surveillance just wasn't working as it's meant to, and they weren't aware that anyone was there at all. I think that's a systemic failure in either case. And the other thing that's really important here is that we only know what was found, right? It really highlights what happens when you just throw technology at a problem as complex as this.

Megan McCarty-Carino (4:07)

You did speak with a Border Patrol officer who was patrolling an area when these new AI cameras went in. And he described some of the problems with the technology. I mean, what kind of stuff came up?

Eileen Guo (4:21)

Yeah, for this story, we spoke to a lot of former Border Patrol agents as well as humanitarian groups that respond to migrant calls of distress. And in terms of the technological issues that came up, there were a whole range of things. You know, there were cases where a Border Patrol agent, because, you know, they're human and they have this experience on the border, they could tell that there was, for example, a raft that was crossing the river. And it was clear to them that it was, you know, a raft of migrants that had, you know, put a tarp over it or something, but an AI might see something like that and not understand what exactly it is. And so it wouldn't be sending out these alerts that it was meant to.