Can AI systems make moral decisions in war?

What the report says

Al Jazeera, in a July 21 report by Hanna Duggal, examined whether artificial intelligence systems can make moral decisions as militaries increasingly integrate them into targeting and battlefield command. The article says AI is already affecting conflicts through interceptor drones in Ukraine, AI-assisted target generation used by Israel in Gaza, and Palantir’s Maven Smart System in the US-Israel war on Iran.

The report focuses on Project Maven, a Pentagon AI programme launched in 2017 to process large volumes of surveillance data. Al Jazeera says Maven now draws from sources including satellites, drone video, radar imagery, signals intelligence, field reports and social media geolocation, helping move targets through the military “kill chain” from identification to assessment. It also reports that Anthropic’s Claude has been integrated through Palantir’s platform, allowing operators to query battlefield data in ordinary language.

A central concern raised in the article is whether faster machine-assisted targeting leaves enough room for meaningful human judgment. Al Jazeera cites US military officials who say humans will retain final authority over firing decisions, while noting that compressed timelines can turn processes that once took hours or days into seconds. The report cites Elke Schwarz, a political theory professor at Queen Mary University London, who argues that moral and legal deliberation require time and cannot be reduced to computational outputs.

The broader significance is the expanding role of AI in warfare as military spending and defence technology investment rise. Al Jazeera frames the debate as moving beyond accuracy: even if AI can sort data and recommend actions quickly, the unresolved question is whether it can understand human vulnerability, suffering and the ethical weight of lethal force.