The US Spent Billions on Border Surveillance: Why AI Towers Failed to Prevent Migrant Deaths

A joint investigation titled *Dying on Camera* by *MIT Technology Review* and *Times of San Diego* reveals significant shortcomings in the billions spent on high-tech border surveillance along the southern US border. The report highlights the case of José Morales Bernal, who crossed into southern New Mexico on April 8, 2024. During his transit, Morales moved within the coverage zone of three newly deployed surveillance towers operated by US Customs and Border Protection (CBP). Built by defense tech firm Anduril, the towers feature automated computer vision and AI designed to identify individuals, stream video to control centers, and alert agents' mobile devices. Although designed to automate monitoring across the 1,951-mile border and enable rapid medical or operational interventions, the automated infrastructure failed to trigger an effective response. Morales was first seen by local landfill workers rather than patrol units. Hours later, after he collapsed in the desert heat, the workers alerted CBP, whose responding personnel appeared unaware that he had spent hours in the immediate vicinity of the AI surveillance systems.