Market & Regulation
MIT investigation exposes AI border surveillance failures
MIT Technology Review AIhigh confidence
Why it matters
- →AI surveillance systems are failing in real-world safety-critical scenarios despite heavy investment.
- →There is a lack of accountability mechanisms to measure if border AI actually improves outcomes.
- →Data collection on migrant deaths near surveillance infrastructure is fragmented and unreliable.
MIT Technology Review has published an investigation into systemic failures within the US government's 'virtual wall' of AI surveillance towers along the border with Mexico. The report identifies cases where migrants died near Anduril-built towers that failed to detect them or where alerts were ignored by agents. Despite a $1 billion federal investment to triple the network by 2034, Customs and Border Protection lacks a comprehensive audit of tower effectiveness or reliable data on whether the technology leads to apprehensions. The article outlines four recommendations for CBP, including mandatory audits and better death tracking protocols.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1MIT Develops Method to Detect Harmful AI Models
MIT News AI · July 13, 2026 · Background
- 2AI Guardrails Hinder Cybersecurity Research Efforts
TechCrunch AI · July 24, 2026 · Background
- 3MIT Develops AI Method for Safety-Critical Tasks
MIT News AI · September 14, 2026 · Background
What happened next
- 4MIT uses AI to map border deaths from police reports
MIT Technology Review AI · September 21, 2026 · Related
- 5Anthropic AI model submits false homicide tip to police
TechCrunch AI · October 9, 2026 · Background
- 6AI Guardrails Fail Rapidly Against Terrorist Prompts
WIRED AI · October 14, 2026 · Related
- 7UK Home Office Faces Probe Over OpenAI Asylum Tools
WIRED AI · October 27, 2026 · Background
- 8AI agents lack real-time observability standards
MIT Technology Review AI · October 29, 2026 · Background
Follow this story
Open the full story →MIT investigation exposes AI border surveillance failures
2 developments
- Sep 21 · MIT Technology Review AI
MIT investigation exposes AI border surveillance failures (This article) - Sep 21 · MIT Technology Review AI
AI border towers fail to prevent migrant deaths
More from MIT Technology Review AI
More in Market & Regulation
© TechCrunch AIMarket & Regulationbusiness
Nebius delays data centers, cuts revenue forecast
Nebius is facing the harsh reality of scaling AI infrastructure: even with $40 billion in planned spending and massive prepayments from Microsoft and Meta, physical construction delays are derailing timelines. A key New Jersey site is months behind schedule due to regulatory fines and generator issues, forcing a shift in delivery dates to 2027. This isn't just a logistical hiccup; it signals that the 'neocloud' model of financing builds with customer cash is fragile when third-party operators fail. The market reacted swiftly, dropping shares by 17% after Nebius cut its 2026 revenue forecast, proving that hype cannot override concrete capacity.
TechCrunch AI
© TechCrunch AIMarket & Regulationagents
United Airlines blocks AI agents from booking flights
United Airlines is actively blocking third-party AI agents from accessing its booking systems, marking a significant shift in how legacy infrastructure handles autonomous software. The airline cites cybersecurity risks from shared credentials and potential violations of federal aviation rules requiring human check-ins. This move aligns with similar warnings from major banks like Bank of America, signaling that critical industries are drawing hard lines around where human oversight is legally required. It effectively kills the dream of fully autonomous travel booking for now, forcing a re-evaluation of agent integration in regulated sectors.
TechCrunch AI
© TechCrunch AIMarket & Regulationbusiness
SoftBank's $7B loss caps AI natural gas ambitions
SoftBank’s massive $30 billion bet on OpenAI and a parallel $33 billion push into natural gas power infrastructure has hit a wall, resulting in a $7.1 billion quarterly loss. The core issue wasn't just bad luck; it was a strategic overreach where SoftBank paid non-refundable deposits for 13 gigawatts of turbines—50% more than needed—to secure supply in a constrained market. This move effectively acted as expensive price discovery, revealing that the anticipated immediate demand for standalone gas plants is slower to materialize than predicted. As data centers realize grid connections paired with battery storage are more viable, SoftBank’s strategy to corner the energy market has stalled, signaling a potential ceiling for AI's reliance on natural gas.
TechCrunch AI