AI system used in air strike that killed Iranian leader Khamenei
Reportedly, an air strike in Tehran on February 28, 2026 killed Iranian Supreme Leader Khamenei and over 40 senior officials. According to reports, the strike employed AI-driven systems including satellite imagery, autonomous algorithms, and real-time intelligence analysis to overcome Iranian air defenses and conduct the operation.
What happened
On February 28, 2026, an air strike in Tehran killed Iranian Supreme Leader Khamenei and over 40 senior officials. According to reports, the strike was characterized as an 'algorithm-dominated decapitation operation' orchestrated by an AI-driven combat system integrating satellite data, real-time algorithms, and autonomous systems.
The operation reportedly used SpaceX's Starshield satellite constellation comprising 480 hardened satellites with 200 Gbps laser links to establish uninterrupted communications networks that breached Iranian electronic countermeasures. Iranian S-300PMU2 air defense systems and domestically-produced Bavar-373 systems reportedly failed to respond. The system transmitted petabyte-scale battlefield data to rear analysis engines within seconds using UAT-222 terminals carried by U.S. special forces.
Palantir's Gotham platform and MetaConstellation system served as the 'battlefield brain,' integrating multi-source data to form a dynamic digital battlefield picture. The platform's algorithms were reportedly deeply integrated with Anthropic's Claude large language model, compressing intelligence assessment cycles from days to hours. Claude reportedly analyzed official media, social media information, simulated traffic flow, and security deployments to predict Khamenei's meeting location coordinates.
Reported impact
Over 40 senior Iranian officials were killed in the strike alongside Supreme Leader Khamenei. The operation demonstrated AI-enabled penetration of advanced air defense systems.
The incident prompted analysis that AI military applications would accelerate China's technology autonomy efforts and raised international concerns about the militarization of AI systems. Anthropic's CEO had previously expressed ethical concerns about unrestricted government use of AI technology, autonomous weapons, and AI-assisted mass surveillance systems.
Index analysis
The operation reportedly demonstrated autonomous systems making real-time tactical decisions within an integrated command structure, utilizing satellite communications, algorithmic target prediction, and multi-platform data fusion to overcome air defenses and deliver kinetic strikes.
Anthropic had previously refused unlimited military access to its technology, and its CEO publicly objected to government unrestricted AI use. Despite this, the Trump administration announced plans to stop using Anthropic while the Pentagon continued using Claude because Palantir systems had already deeply integrated the model and could not be replaced short-term.
Governance controls
The U.S. Department of Defense coordinated with multiple technology companies including Anthropic, Palantir Technologies, SpaceX, and others to deploy integrated AI systems for military operations. The Pentagon continued using Claude AI despite Trump administration orders to discontinue Anthropic relationships.
China's Foreign Ministry Deputy Director of Arms Control emphasized adherence to 'human-centered military AI' principles, stating that countries should maintain cautious and responsible conduct, uphold human-centered principles, follow intelligent good practices, implement agile governance, and advance multilateralism when using such technologies.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.