The Pentagon is rewriting the rules of war: AI is granted the right to aim weapons, but under human supervision.
The U.S. Department of Defense has approved a new, significantly expanded doctrine for the use of artificial intelligence on the battlefield. The document, signed back in April, is not classified, but its content remained out of public access until now. This represents a fundamental change in the role of algorithms in military operations.
The key innovation is the official recognition of "combat systems in which AI initiates actions under human monitoring." This means that neural networks gain the right not just to analyze data, but to independently launch processes leading to the use of force. The speed of modern conflict and the technological progress of potential adversaries, as stated in the document, compel the U.S. to consider creating "fully autonomous systems."
The main goal of the doctrine revision is a radical reduction of the "sensor-to-target kill chain." AI must now not only process intelligence data but also correlate information from dozens of platforms in real time, creating a unified operational picture. Commanders are instructed to more actively use algorithms for risk analysis, decision-making, and selecting priority targets.
Notably, the document addresses "reducing harm to civilian populations." However, it immediately follows with a warning about the "serious moral and legal dilemmas" that excessive reliance on machines creates. The developers emphasize: automation does not replace human thinking and proactive communication.
The official Pentagon position, voiced by a department representative, boils down to the fact that humans always remain informed of key operational decisions. Allegedly, existing AI technologies do not allow for autonomous target selection or strike execution. Commanders, they say, retain full responsibility for every decision.
Almost simultaneously with the article's publication, the launch of an "AI agent network" for battle management and target selection was announced. As stated by the Pentagon's Chief Digital and AI Officer, Cameron Stanley, this network will provide commanders with rapid access to quality information, keeping human judgment at the center of the process.
Recall that back in 2023, the UN Secretary-General called for a complete ban on lethal autonomous systems, calling them "morally repugnant." However, judging by the dynamics, military departments are moving in the exact opposite direction. Against the backdrop of a reported 1775% increase in AI adoption at the Pentagon, this step looks like a logical but extremely risky stage of the arms race.
Analyst's comment: This document is not just an update of instructions, but the removal of the last formal barriers before transferring the right to violence to machines. Statements about "human control" are more of a ritual formula designed to reassure the public. In a real battle, where seconds count, "monitoring" will turn into a mere formality. The crypto industry should closely monitor these processes: the decentralization and verification technologies we are developing could become the only tool for creating transparent and ethical weapons management systems.