AI warfare in Middle East leaves digital evidence trail
TL;DR:
- AI systems including Palantir’s Maven and Israel’s Lavender are dramatically accelerating military targeting in Iran and Gaza, with US forces hitting over 2,000 targets in four days.
- The same AI kill chains that speed up strikes also create auditable digital records of every targeting decision — making war crimes harder to conceal.
- Palantir signed a major contract with the UK Ministry of Defence last December, making the ethical and legal implications of these systems directly relevant to British defence policy.
Artificial intelligence is reshaping how wars are fought in the Middle East, compressing the targeting cycle from hours to minutes while simultaneously creating digital records that could be used as evidence in future legal proceedings.
Two AI systems, two approaches
The US military’s use of Maven — an AI system developed with Palantir and sold to the Pentagon — has enabled more than 2,000 strikes in Iran over just four days. Maven processes intelligence across multiple formats to identify targets, match them with appropriate weapons, and estimate civilian casualties before human commanders authorise strikes.
Israel’s own systems operate differently. Lavender, first revealed by Israeli magazine +972, identifies individual suspected militants, while a companion system called Gospel flags buildings likely used by armed groups. Combined, these tools contributed to the targeting decisions behind over 70,000 deaths in Gaza according to reported figures, and are believed to be in use during current operations in Iran.
The accountability paradox
What makes AI-enabled warfare distinct is not just its speed but its traceability. Unlike conventional military operations, AI kill chains generate comprehensive logs of targeting rationale, weapons selection and damage estimates. As one insider quoted in the original reporting put it: “They’re available for instant audit.”
This creates a tension that defence establishments are still grappling with. The +972 investigation alleged that Israeli forces programmed acceptable civilian casualty ratios — up to 15-20 civilians per junior Hamas operative — into their AI systems. The IDF denies these claims and all allegations of deliberately targeting civilians.
UK implications
Palantir’s December deal with the UK Ministry of Defence means British forces are now part of this evolving picture. As NATO also moves to adopt Palantir’s technology, questions about AI targeting rules, acceptable casualty thresholds and accountability frameworks are no longer abstract policy debates for UK defence officials — they are operational decisions that will need clear governance structures.
Looking forward
The dual nature of AI warfare — faster killing alongside harder-to-hide evidence — may ultimately reshape how international law applies to armed conflict. For UK policymakers and defence contractors, establishing transparent rules for AI-assisted targeting before these systems see wider deployment will be easier than retrofitting accountability after the fact.