AI border surveillance towers failed to prevent migrant deaths in plain sight of cameras, investigation finds

A MIT Technology Review investigation finds that a 30-year-old migrant died within view of an AI-equipped surveillance tower near San Diego, part of a pattern of 138+ deaths near California towers over four years, with no CBP audit of the system's…

An MIT Technology Review investigation into the US border 'virtual wall' reveals that AI-enabled surveillance towers have failed to prevent migrant deaths, even when migrants die within the cameras' line of sight. Graciela Gómez Hernández, a 30-year-old migrant, died within plain sight of a surveillance tower near San Diego. The investigation found over 138 migrant deaths in the past four years near California towers, and topographical analysis shows over half of California deaths occurred within camera line-of-sight. Failures include both towers failing to detect people and agents failing to respond to alerts. Despite the US spending billions on border surveillance, CBP has never conducted a comprehensive audit of tower effectiveness, and federal spending on the virtual wall is set to increase significantly, with plans to spend $1 billion expanding the system.

• Graciela Gómez Hernández, a 30-year-old migrant, died within plain sight of an AI-equipped surveillance tower near San Diego.
• Over 138 migrant deaths occurred in the past four years near California surveillance towers.
• Topographical analysis shows over half of migrant deaths in California were within the cameras' line of sight.
• The towers use AI to detect and track movement, but migrants continue to die of exposure in plain sight.
• Failures include both towers failing to detect people and agents failing to respond to alerts.
• CBP has never conducted a comprehensive audit of tower effectiveness.
• The US has spent billions on border surveillance, with plans to spend $1 billion expanding the 'virtual wall' system.