Technology & OSINT — 2026-03-22

Wired Exposes Palantir's Use of Claude AI in Iran War Targeting; Congressional Oversight Demands Mount

Reporting reveals that Palantir's Maven Smart System incorporates Anthropic's Claude large language model to scan and summarize intelligence documents feeding the Iran targeting pipeline. The system condenses hours of analyst work into minutes, processing data from drones, satellites, and radar. However, concerns mount that the human in the loop has become a 'rubber stamp' for machine-made targeting recommendations. The U.S. military confirmed using 'advanced AI tools' in Iran operations and used Claude-assisted targeting to strike over 1,000 targets in the first 24 hours of Operation Epic Fury.

Analysis
Claude's integration into the Maven targeting pipeline, combined with the Minab probe findings and Maven's program of record formalization, creates a three-way accountability question: Anthropic's acceptable use policies, the Pentagon's AI governance framework, and Congressional oversight authority over AI-assisted targeting. The 121-Democrat letter deadline passed unanswered.
3 sources
  1. Wired Exposes How Palantir Uses Claude AI in Iran War Targeting - The Dupree Report / Wired
  2. US used 'Claude' to strike over 1000 targets in first 24 hours of war - Responsible Statecraft
  3. Speeding Up the Kill Chain: Pentagon Bombs Thousands of Targets in Iran Using Palantir AI - Democracy Now

View in full brief →

UNCLASSIFIED // OPEN SOURCE