Pentagon Revises Targeting Doctrine to Let AI Initiate Actions With Human Monitoring
Updated
Updated · Bloomberg · Jun 25
Pentagon Revises Targeting Doctrine to Let AI Initiate Actions With Human Monitoring
3 articles · Updated · Bloomberg · Jun 25
Summary
April-approved Pentagon guidance quietly shifts US targeting doctrine from humans initiating battlefield actions to AI systems that can initiate them under human monitoring.
The revision broadens AI’s role in critical wartime decisions by moving beyond current “human in the loop” practice, where a person must start the action.
The change signals a wider future role for artificial intelligence in how the US military selects and engages targets in combat.
As AI takes the lead in targeting, can human oversight prevent a 'fog of AI war'?
Is the Pentagon's new doctrine launching an irreversible global autonomous arms race?
The Pentagon’s 2026 AI-First Doctrine: Speed, Risk, and the Global Race for Autonomous Weapons
Overview
In 2026, the Pentagon made a major policy shift by adopting an 'AI-First' doctrine, driven by the belief that delaying AI integration poses a serious strategic risk to U.S. military readiness and global competitiveness. Defense Secretary Pete Hegseth emphasized that this move is essential to equip American forces with advanced technologies and maintain an edge over adversaries at wartime speed. Central to this doctrine is the idea that speed is decisive in military AI, with leaders urging the rapid adoption and measurement of AI learning and operational cycles. This marks a new era where fast, AI-driven decision-making is seen as critical for future military success.