AI’s Real-World Killings Demand Action, Not Alarm

AI’s Real-World Killings Demand Action, Not Alarm

AI’s existential warnings from tech billionaires make headlines, but the technology’s most devastating impacts are already landing on civilian populations. As documented in a recent Guardian piece by Granate Kim and Mohamed Hussein, Israel’s military used AI to generate 12,000 bombing targets in Gaza within the first month of the war, with minimal human oversight. The U.S. military’s AI-assisted airstrike on an Iranian elementary school killed 156 people — a targeting mistake fed by faulty data. Meanwhile, the Department of Homeland Security deploys Microsoft’s Azure to monitor and deport immigrants, and police departments nationwide use facial recognition that has wrongfully arrested Black and Latino Americans.

These are not hypothetical future harms — they are happening now. While the public debate fixates on whether superintelligent AI could kill us all by 2030, states and corporations are already deploying AI as a weapon of control and destruction. The same companies that sign open letters about existential risk — Microsoft, OpenAI, Google — have been building the infrastructure for these systems. Employee movements like No Azure for Apartheid have pushed back, but the contracts remain.

This reality directly affects individual freedom. When AI surveillance becomes routine, when autonomous weapons make life-or-death decisions without accountability, every civilian loses agency. The power asymmetry between governments and individuals grows steeper. As noted in Cogtropolis coverage, Anthropic clashed with the Pentagon over domestic surveillance, and Clearview AI offered free trials to governments — both examples of the creeping militarization and normalization of AI surveillance. The EU’s AI labeling rules are a step toward transparency, but they do nothing to stop bombs or deportations.

The path forward requires more than alarmist warnings. It demands concrete bans on AI in military targeting and mass surveillance, third-party safety audits, and robust enforcement — not voluntary corporate promises. Individuals can support employee-led movements, push for local surveillance oversight, and choose technologies that respect privacy. The future is already here; it’s time to act like it.

Source: The Guardian

This article is newsworthy because it moves the AI risk debate from hypothetical future scenarios to documented present harms in Gaza, Iran, and U.S. neighborhoods — directly relevant to Cogtropolis readers who care about surveillance, state power, and individual freedom. The timing aligns with ongoing policy debates (EU AI Act, U.S. Frontier Act) and recent employee activism at major AI companies.

Source: Original article