AI weapons systems are already here. Algorithms must not decide who lives and dies
Israel’s conduct in Gaza shows the danger of using AI in warfare. Humans must remain at the center of wartime decisions
TL;DR
- Autonomous AI weapons, or killer robots, are a growing threat, with AI already used for targeting in conflicts like the one in Gaza.
- International negotiations to ban autonomous AI weapons are underway, but progress is hindered by resistance to provisions requiring meaningful human review of AI-generated targets.
- Israel's use of AI systems like "Lavender" and "Where's Daddy?" to identify targets and their residences in Gaza has led to a high number of civilian casualties, with human oversight described as a nominal "rubber stamp".
- The "banality of evil" is exacerbated by AI's feigned omniscience, which can lead to complacency and deference to algorithmic judgments, magnifying human biases and devaluing civilian life.
- Ensuring meaningful human control requires intense questioning of AI's output and accountability for commanders, rather than mere nominal endorsement of AI decisions.