Pentagon Maps Autonomous Warfare Command Vision, Pushing Drone Dollars to Frontline Units

Project Agincourt, the Department of War effort charged with standing up a four-star Autonomous Warfare Command (AUTOWARCOM) by October 1, 2027, has told industry and the joint force that combat results, not paperwork, will decide which unmanned systems the new command scales. In an open letter dated October 19, 2026, and released October 21, chief executive officer Owen West and chief operating officer Senior Chief Max Strasiser set out a combat-first model that pushes procurement decisions and dollars directly to frontline warfighters.

The letter follows the September 30, 2026, memorandum in which Secretary of War Pete Hegseth created Project Agincourt. That memo assigned two missions: build the pathway to a functional combatant command that will lead the joint force in unmanned autonomous warfare, and continue the work of the Direct Reporting Portfolio Manager for Unmanned Systems (DRPM-UxS), the office Hegseth created on June 29, 2026, as the department’s single joint integrator for unmanned and autonomous programs. The memo gave the team 30 days to deliver a plan of action and milestones with the Chairman of the Joint Chiefs of Staff.

How the Autonomous Warfare Command Will Measure Success

West and Strasiser framed the command around one output: speed of adaptation. Success, they wrote, will be measured by how quickly an operational problem receives a fielded solution that can scale. The stated objective is not a particular drone but a system built to adapt at scale.

“Since the Cold War, American power projection has rested on a small number of sophisticated, expensive weapons,” they wrote. “That model has been disrupted. The U.S. can build a new generational advantage by adapting at the edge faster than adversaries by leveraging two strengths: our lead in super intelligence, and an unmatched [noncommissioned officer] corps.” Super intelligence is the administration’s term for advanced artificial intelligence.

“Urgency does not excuse imprecision,” the pair wrote. “We will work analytically with the Joint Staff, the services and, above all, the operational units that will serve as our incubators. Evidence will decide outcomes. Competitions, exercises and combat performance will determine which systems scale, and which are cut.”

The letter also imposes a fiscal test. Citing the “power of the purse,” West and Strasiser wrote that every auditable dollar should reduce either cost-per-kill or risk-to-force, the yardsticks meant to flip the cost exchange that cheap drones impose on expensive weapons.

Warfighting Acquisition and Theater-Specific Drone Portfolios

The second Agincourt mission is a new acquisition model Hegseth calls warfighting acquisition. It pairs autonomy and robotics companies directly with operational units, moving decisions and funding toward combatant commands and frontline formations rather than Pentagon program offices. Hegseth said this month that $1 billion in Drone Dominance funding has already gone straight to operators.

The letter sets a hard readiness standard: no system will be declared ready on paper, and combatant commanders must be able to employ it in combat. It also rejects a uniform joint inventory: the best Pacific portfolio is not the best European one, so each theater builds its own mix while the services are strengthened rather than displaced. The authors also pledge not to replace one bureaucracy with another.

The DRPM-UxS holds directive authority over Group 1 to 3 unmanned aircraft, unmanned surface and ground vehicles, autonomy and swarming software, counter-unmanned systems and unmanned logistics. It reports to Deputy Secretary of War Stephen Feinberg and can direct service contracting activities. The services are aligning: on October 1, Acting Army Secretary Adam Telle directed the creation of an Army Futures and Autonomous Systems Command and a Portfolio Acquisition Executive for Autonomy.

Interwar Lessons and the Congressional Path to AUTOWARCOM

The authors anchor their case in interwar history. In 1928, Secretary of War Dwight Davis ordered an experimental mechanized force assembled at Camp Meade, Maryland. The experiment was not sustained, and the Army created its Armored Force only in July 1940, after German armored operations in France. The letter casts low-cost drones as a comparable revolution and presents AUTOWARCOM as the decision to institutionalize change before war forces it.

The command Hegseth described resembles U.S. Special Operations Command (SOCOM): four-star leadership, service-like authorities, its own manpower and budget, and career paths for officers and enlisted personnel. The president can establish a combatant command under existing law, but SOCOM-style budget and acquisition authorities likely require legislation, and the commander requires Senate confirmation. Section 917 of the Senate Armed Services Committee’s fiscal 2027 National Defense Authorization Act would authorize a similar Robotic and Autonomous Systems Command with its own budget line and operational test authority, though with narrower acquisition powers; the provision is Senate-only.

Funding is the binding constraint. The fiscal 2027 request seeks $54.6 billion for the Defense Autonomous Warfare Group, but only about $1 billion sits in the base budget request; the rest depends on reconciliation legislation Congress has not passed.

Analysis: The letter’s most consequential choice is to make the operational unit, not the program office, the decision point for scaling. Tying spending to cost-per-kill and risk-to-force gives the command an auditable standard most acquisition programs lack, and theater-specific portfolios recognize that a maritime Pacific fight and a land-centric European fight demand different autonomy mixes. The friction is structural: a four-star command with budget authority will compete with the services for money, people and requirements, and David Deptula, dean of the Mitchell Institute for Aerospace Studies, has warned that centralization could make autonomy less responsive to the commanders who need it. Whether Agincourt avoids building a new bureaucracy depends on authorities Congress has not yet granted.

Next come the Agincourt plan of action due under the memo’s 30-day deadline and a four-star commander, whom Hegseth has said he will name soon. House and Senate negotiators must still reconcile their defense authorization bills, and the letter does not say when testing with incubator units will begin. “We are on the clock,” West and Strasiser wrote, with less than 12 months left before the October 1, 2027, target.