AINEWS
Defense Secretary Pete Hegseth announced the rare four star command at Quantico, with establishment contingent on Congress and targeted for October 1, 2027.
Defense Secretary Pete Hegseth on Wednesday announced a new Pentagon command, the Autonomous Warfare Command, dedicated to building cheap, AI-driven drones and autonomous systems across the US military (Tribune report). The reorganization is rare: a new combatant command is the top tier of Pentagon structure, and the last one stood up was US Space Command in 2019 (Pentagon announcement).
Hegseth described the move as a doctrinal shift, not a procurement adjustment. He told a "State of the Force" audience at Quantico, Virginia, that "cheap compute, superintelligence, and advanced commercial manufacturing" now let the United States field "low-cost, high-precision strike" at scale (The War Zone). The unit of account is changing: from the price of the platform that carries a weapon to the price of the software that finds the target.
A four-star officer will lead it, and Hegseth described service-like authorities and a new acquisition model that puts operators and entrepreneurs in the same rapid adaptation cycle, with "decisions and dollars" closer to frontline units (The War Zone). The Pentagon's official account names the pathway: Project Agincourt, run by the direct reporting portfolio manager for unmanned systems, will establish the command, execute the existing unmanned-systems mission, and prototype the new acquisition model (Pentagon announcement). Each military department will also field new military occupational specialties for autonomy work.
The timing is contingent. The Pentagon memo reported by The War Zone ties establishment to congressional approval and targets October 1, 2027. Until then, existing service-level and Pentagon drone initiatives are the baseline, not an absence of autonomous capability. The new command is a proposal layered on top of them, with service integration, congressional support, and budgeting still open (The War Zone).
The Pentagon already has eleven combatant commands and a thicket of drone and autonomy programs; standing up a four-star command over them is mostly a way to make autonomy someone's day job. That is a fair read, but it understates what Hegseth actually said. The acquisition-model change, with operators and entrepreneurs in one rapid cycle, is a structural change to how the Pentagon decides what to buy and how fast. That is not a reorg chart; it is a new way to spend the budget.
The trade Hegseth is making is visible in the expert reaction, not in the announcement. Security researchers and arms-control specialists in both the United States and China have pressed for guardrails on AI in military systems, including in nuclear command and control, on the grounds that cheap autonomous mass compresses decision time in ways current doctrine does not handle. The Pentagon's "superintelligence" and "low-cost, high-precision strike" language sits directly on that fault line. Hegseth's argument is that the United States cannot afford to slow its own adoption to wait for international rules, because the alternative is ceding mass to a peer competitor.
The next concrete milestone is congressional posture. The October 1, 2027 target gives lawmakers roughly a year to decide whether to fund, modify, or block the new command. The new occupational specialties and the Project Agincourt acquisition prototype hit the budget cycle first. If Congress funds them and the four-star command, the shift is real; if Congress narrows the request, the command becomes another layer over existing programs.