The United States has spent billions of dollars on surveillance towers along the border, building what officials describe as a virtual wall. But an investigation into the system found a troubling gap: people have died near these towers without the deaths being fully recognized in the technology’s record of border activity.
More than 18 people died near surveillance towers in one small stretch of desert in 2024. Across the border, the investigation identified more than 1,050 migrant deaths near nearly 600 towers. Those numbers raise a basic question about what the towers are designed to measure, and what they fail to see.
A growing system with a deadly blind spot
The US government is set to spend $1 billion to triple the size of the virtual wall by 2034. That planned expansion comes as Border Patrol estimates that more than 10,000 people have died during crossings since 2000.
The surveillance network includes towers intended to monitor border areas, but the investigation found systemic failures in tracking deaths near them. The technology may help identify movement across remote land, yet its records do not provide a complete picture of what happens to people who cross, become stranded, or die nearby.
The problem is not limited to one agency or one missing record. The investigation examined cases dating back to 2015 and showed how difficult it can be to connect a death, a suspected border crossing, and a surveillance tower. The result is a record that can miss people even when their remains are found close to the system’s equipment.
Sarah Silva, a New Mexico state representative, described the concern in direct terms: “If I were the federal government, I would do an audit, because this feels like it’s not working.”
Building a clearer record of border deaths
MIT Technology Review and Times of San Diego conducted a 15-month investigation into border deaths and surveillance failures. The work analyzed cases from 2015 through April 2026 and drew on data from more than a dozen agencies and organizations, including No More Deaths, Humane Borders, and the Electronic Frontier Foundation.
Records requests went to 17 counties in Texas, and usable records came back from 14 counties. The investigation reviewed more than 4,000 pages of police reports, including handwritten documents, and assembled more than 1,500 cases from Texas records, more than 1,700 cases from Humane Borders, and over 1,500 cases from other sources.
Researchers treated a case as eligible for analysis only when they could confirm three points: the person was believed to have been crossing the border, the place where the remains were found, and roughly when the person died. That standard matters because border deaths are recorded across different agencies and systems, with no single record covering every case.
The location data also varied in precision. Some cases had GPS coordinates accurate to within 300 feet, while others placed the remains within a quarter-mile. Those details allowed the investigation to compare deaths with the locations of surveillance towers without claiming that every death near a tower resulted from a tower failure.
What the expansion must answer
The planned growth of the virtual wall puts pressure on the US government, Border Patrol, and CBP to explain how the system will handle deaths as well as suspected crossings. A larger network will not solve the problem if agencies do not track the people who disappear from its view.
The first issue is accountability. The total of more than 1,050 deaths near towers shows why an audit matters before the government triples the system’s size. An audit could examine whether tower records, police reports, medical examiner files, and border data are being compared in a way that reveals missed cases.
The second issue is completeness. Border Patrol’s estimate of more than 10,000 deaths since 2000 is already a warning that the human cost of crossings extends beyond the incidents captured by surveillance technology. Records from counties, nonprofit groups, and other agencies each cover part of the picture, but the investigation had to combine them to identify patterns.
The third issue is distance. A tower can stand near a death without providing a clear explanation of what happened. Some cases in the investigation had precise coordinates, while others were known only within a quarter-mile, so any review needs to preserve that difference instead of treating every location as exact.
The fourth issue is whether spending more will produce better results. The US has invested billions in surveillance towers and plans to spend another $1 billion to triple the virtual wall by 2034. The deaths documented near nearly 600 towers show that expansion should be judged not only by how much territory the system covers, but also by whether it produces a more complete record of people who die during crossings.
The virtual wall was built to watch the border. The investigation shows that watching movement is not the same as understanding its consequences. Before the system grows, the government faces a clear test: determine why more than 1,050 deaths near towers were not fully reflected in its tracking systems, then make the record impossible to overlook.
Based on
