Story
September 22, 2026
Border Surveillance’s Billion-Dollar Expansion Faces a Deadly Accountability Gap
The investigation’s central argument is not that surveillance alone can solve a humanitarian crisis, but that a rapidly expanding system must be held accountable when people die within its reach. Its proposed remedy is rigorous auditing, better death records and automatic reviews of whether technology and agents failed to act.
The U.S. border’s “virtual wall” was built to see what agents could not. But an investigation by MIT Technology Review and Times of San Diego argues that its blind spots have become lethal: migrants have crossed areas watched by advanced towers, then died nearby without being found for weeks or months.
The accountability problem stretches back years. Border Patrol began using the tower types examined in the investigation in 2015, while its internal Border Safety Initiative Tracking System has existed since 2007. Yet the reporting found no comprehensive agency audit matching deaths to the surveillance network. The journalists mapped nearly 600 towers and identified more than 1,050 deaths nearby — a likely undercount, because CBP does not publish a complete tower map and the government says about 800 operate today.
The same weakness has surfaced in official reviews. In 2014, the Government Accountability Office urged Border Patrol to record when technology helped produce apprehensions. By 2017, the GAO found the resulting data unreliable, including cases in which Texas agents credited towers that did not exist in the state. CBP completed an effectiveness analysis in 2022, but its results have not been made public.
By 2024, the human toll was visible even in a small area of desert near New Mexico state representative Sarah Silva’s district, where more than 18 people died near towers. Shown the findings, Silva said, “If I were the federal government, I would do an audit, because this feels like it’s not working.”1
The investigation says the response should go beyond counting bodies: CBP should standardize death records, review every death occurring within surveillance range, preserve footage that is otherwise typically deleted after 30 days, and examine whether a tower detected a person and whether agents responded.
That pressure is mounting as the virtual wall grows. The U.S. is expected to spend $1 billion to triple the network by 2034 — potentially expanding a system whose effectiveness, and failures, remain only partly visible.