MIT Technology Review's investigation reveals systemic flaws in the US-Mexico border's 'virtual wall' of AI-enabled surveillance towers. The report found over 1,050 migrant deaths near towers, with cases of people walking undetected and dying unnoticed for weeks due to broken towers, failed algorithms, and unresponsive agents. The investigation, part of the 'Dying on Camera' collaboration with Times of San Diego, analyzed nearly 600 towers but estimates the government operates 800, suggesting the death toll is an undercount.

The US plans to spend $1 billion to triple the virtual wall by 2034, with defense company Anduril building the latest towers and securing major contracts. To address these failures, the article proposes four actions: first, conduct a comprehensive audit of deaths near the virtual wall to accurately track tower locations, installation times, and mortality patterns—something CBP has never done despite having the capability since 2015. Second, fix the broken system for tracking migrant deaths, as no national agency collects complete records and Border Patrol's internal BSITS system lacks standardized protocols.

Third, improve agent response to alerts from surveillance systems to ensure timely intervention. Fourth, increase transparency by publishing a comprehensive map of the virtual wall and sharing audit findings publicly to enable accountability and informed policy decisions.