AutoWarCom and the Acquisition Question: What 'Service-Like Authorities' Actually Mean for Autonomous Systems

BLUF — Bottom Line Up Front

On September 30, 2026, Secretary of War Hegseth announced the Autonomous Warfare Command (AutoWarCom) at Marine Corps Base Quantico — the proposed 12th combatant command, organized around drones, AI, and C2 integration, with a target operational date of October 1, 2027 pending Congressional authorization (Inside Unmanned Systems). The practical consequence is structural: if AutoWarCom receives the "service-like authorities" Hegseth described, it would join SOCOM and, more narrowly, CYBERCOM as the only combatant commands holding their own acquisition and budget authority, and the first organized around a capability set every service is also buying — potentially reshaping how autonomous systems are funded and fielded. The reader's next decision is whether to treat AutoWarCom as a future buyer of record or as a coordination body, and the answer should drive near-term positioning and relationship strategy over the next twelve months.

What AutoWarCom Is — and What It Isn't Yet

Combatant commands fight wars. They don't buy equipment. INDOPACOM, EUCOM, CENTCOM, and SOCOM each organize warfighters and execute assigned missions. The military services — Army, Navy, Air Force, Marine Corps, Space Force — own acquisition: they run program executive offices (PEOs), manage development contracts, and receive appropriated funds through service-specific budget titles in the PPBE process.

AutoWarCom is being described differently. Hegseth's September 30 State of the Force address described the command as carrying "service-like authorities." Those three words are doing substantial analytical work. In the context of DoD organizational law, "service-like authorities" most plausibly means budget authority of its own — an appropriation line — and the associated program management infrastructure. That combination is how services buy things; it isn't how commands operate.

There are two exceptions to this pattern. SOCOM is the full one: 10 U.S.C. § 167 gives its commander the functions of an agency head under chapter 137, enabling it to directly procure specialized equipment for special operations forces without routing everything through the Army, Navy, or Air Force. SOCOM maintains its own PEOs and manages its own programs. That structural independence enables faster movement on SOF-peculiar equipment — for common equipment, SOCOM remains subject to the same statutory acquisition requirements as the services — and it's the model AutoWarCom appears to be drawing from. U.S. Cyber Command is a partial exception: under 10 U.S.C. § 167b its commander controls expenditure of funds for matters covered by section 807 of the FY2016 NDAA and is responsible for "directly controlling and managing the planning, programming, budgeting, and execution of resources to train, equip, operate, and sustain the Cyber Mission Force." Breaking Defense noted that Congress has approved acquisition authorities for the Special Operations and Cyber commands in the past, so Congress has already granted such authority in both full and narrower forms.

If AutoWarCom is to have genuinely equivalent authority, it needs Congressional action comparable to what created Section 167. Hegseth's announcement is a statement of intent, and Congress has only partly acted. The Senate Armed Services Committee's FY2027 NDAA (S. 4784) would permit, not require, a four-star Robotic and Autonomous Systems Combatant Command; DefenseScoop reported that the committee language "encourages the department" to establish it, with special test and evaluation authorities and limited acquisition authorities. As of September 1 the Senate had not brought its bill to the floor, according to CRS. The Agincourt memo, per DroneXL, directs the unmanned-systems portfolio office to go to Congress for the legislation the command needs. Whether the final language looks like the limited authority in the Senate bill or something closer to Section 167 is the most important procedural question.

The Acquisition Problem AutoWarCom Would Solve

To understand why an acquisition-capable AutoWarCom would matter, it helps to trace the specific dysfunction it's meant to address.

The services each develop autonomous platforms independently. Army autonomous ground vehicles, Navy unmanned surface vessels, Air Force collaborative combat aircraft, and Marine Corps UAS programs are acquired through separate service programs, often with incompatible data standards, different link architectures, and non-interoperable command interfaces. The result is a collection of capable platforms that can't readily operate as an integrated force. JADC2 — Joint All-Domain Command and Control — is the operational aspiration; the acquisition structure has worked against it.

The Defense Innovation Unit has partially addressed this through Other Transaction Authority (OTA), which bypasses much of the FAR procurement process. But OTA produces prototypes and limited-production contracts. Transitioning an OTA-originated program into a service program of record — the path to scale, sustainment funding, and eventual fleet-wide fielding — requires a service to adopt it. Services have institutional reasons to resist programs that didn't originate in their own requirements process. The "valley of death" between prototype and program of record has swallowed meaningful autonomous capability.

An acquisition-capable AutoWarCom changes this architecture directly. A command with its own program authority doesn't need a service to adopt a capability. It buys it, tests it, and fields it across the joint force. The interoperability standards are set by the command rather than negotiated across the services. The T&E baseline is common rather than service-specific.

That's the structural argument for service-like authorities. It's also the argument the services will resist.

Project Agincourt: Where the Architecture Is Being Written

Hegseth named Owen West — Director of the Defense Innovation Unit — and Navy SEAL Senior Chief Max Strasiser to lead Project Agincourt, the interim effort standing up AutoWarCom.

The choice of West is not incidental. DIU's OTA mechanism and SOCOM's acquisition model are the two existing organizational templates AutoWarCom could draw from. Pairing the DIU director with a special operations operator brings a user-community perspective into the conceptual design from the start.

Project Agincourt is not a placeholder. Its memo tasks it with the organizational shifts, legislation and career tracks the command needs, with a plan of action and milestones due within 30 days (DroneXL). The design choices made in the coming months — whether AutoWarCom holds its own appropriation, its own PEO, its own T&E authority — will almost certainly shape what Congress is asked to authorize. The window for companies to engage the concept while it is still being written is bounded.

A parallel 120-day study, Project Meridian, is led by Pentagon CTO Emil Michael (DefenseScoop) and co-directed by Elon Musk, Palmer Luckey (Anduril co-founder) and former House Speaker Newt Gingrich, with MITRE selected to lead the panel. It is examining the future of warfare and the weapons and capabilities the military should develop, with findings due to the defense secretary no later than January 28, 2027. The co-directors' existing defense industry relationships — Luckey's in particular — make Meridian a political and commercial dynamics story as much as a technical one. Agincourt is the organizational design; Meridian is a broader scan of future warfare, and no source yet ties it to the AutoWarCom statutory proposal.

What Can Fail

The SOCOM precedent is instructive but not uniformly encouraging. After Congress established SOCOM, the command spent much of its early years in jurisdictional conflict with the services over budget authority, program ownership, and the definition of what "special operations forces" actually needed. The arrangement stabilized — SOCOM now runs effective programs — but the friction was real and, in this analysis, likely persisted for years.

The money is already moving: Inside Unmanned Systems reported that the fiscal 2027 budget request seeks $54.6 billion for the Defense Autonomous Warfare Group, up from about $226 million the year before. AutoWarCom will encounter the same resistance in more concentrated form. Autonomous systems are not a niche capability; they are a central priority for the services. Each service has ongoing programs, acquisition infrastructure, and institutional investment in its own autonomous systems path. An acquisition-capable AutoWarCom competes with all of them simultaneously.

Congressional authorization is the harder constraint. A new appropriation title requires authorization and appropriations legislation to move in sync. The October 1, 2027 target assumes action in the FY2027 or FY2028 NDAA cycle. The FY2027 NDAA is already complicated by the divergence between House and Senate positions on AI targeting and Directive 3000.09 review requirements: the Senate committee adopted a Kelly amendment codifying the 2023 directive's human-oversight policy by an 18-9 vote (KJZZ), and the House passed H.R. 8800 216-212 on July 22 (CRS). Adding an entirely new combatant command with acquisition authority to that debate introduces more negotiating surface, not less.

The credible alternative outcome: Congress authorizes AutoWarCom with coordination authorities only — joint requirements validation, interoperability standards, and shared T&E infrastructure — but no independent budget title or PEO structure. In that version, AutoWarCom sets the standards the services must meet; the services still buy. That outcome preserves the command's strategic framing and genuinely addresses interoperability without directly threatening service acquisition prerogatives. Whether it solves the transition problem is a harder question.

Evidence That Changes This Analysis

Three developments would materially alter the acquisition-authority conclusion:

If the final language falls short of the Senate committee bill. The Senate text offers only limited acquisition authority, so the conference outcome on the Robotic and Autonomous Systems Combatant Command provision is the first signal to watch. The language actually sent to Congress matters more than the speech. Watch for whether the proposal includes a dedicated major force program and budget submission comparable to SOCOM's MFP-11, a PEO organizational structure, and a formal designation analogous to Section 167. Absence of those elements means the command's acquisition implications are limited.

If the FY2027 or FY2028 NDAA imposes substantial constraints on autonomous weapons acquisition. NSPM-11, signed June 5, 2026, directed the Secretary of War to update Directive 3000.09 — the DoD policy governing autonomy in weapon systems — within 90 days, as part of a policy of "eliminating unnecessary barriers to rapid deployment" of AI, while keeping commanders "responsible and accountable." Critics, including Sen. Mark Kelly, read the update as a risk to human-judgment requirements; KJZZ reported that the Kelly amendment adopted by the Senate committee codifies the 2023 Directive 3000.09 and would effectively block any elimination of human involvement in the kill chain. If Congress codifies stricter oversight requirements than the revised directive, those requirements would constrain what AutoWarCom is authorized to acquire regardless of its organizational authority.

If Project Agincourt recommends against independent acquisition authority. If the concept analysis concludes the better model is enhanced OTA authority — a DIU 2.0 with broader mandate — rather than a full acquisition command, the statutory proposal will reflect that. The interim organizational design is the signal to watch.

What This Means for Defense AI and Autonomy Companies Now

Regardless of which model AutoWarCom adopts, the announcement signals institutional commitment to treating autonomous warfare as an integrated domain rather than a set of parallel service programs. Several practical implications follow now, before the command structure is finalized.

Relationship positioning. Owen West is leading the organizational design. DIU relationships are the most direct access point. Project Agincourt's conceptual work is happening now, and industry engagement during this phase is typically possible and consequential. If your company has relevant technology — AI command interfaces, autonomy stacks, cross-domain communication architectures — this is the window in which the requirements are being written.

Standards and interoperability compliance. A command that sets joint requirements — even without direct acquisition authority — produces standards. Alignment with those standards becomes a de facto qualification threshold. AI data format choices, API contract designs, abstraction layer architectures made now will either map to AutoWarCom's eventual standards or require expensive rework. Treat the command's concept of operations documents as technical standards when they appear.

T&E positioning. The command will require a T&E framework for autonomous systems. Existing service T&E organizations have service-specific mandates; the joint T&E function for AutoWarCom programs is an open question. Companies with meaningful autonomous systems experience should be tracking the T&E discussion in Agincourt — both because they can contribute to it and because the resulting framework will define what verification and validation looks like for anything the command acquires.

The twelve months before AutoWarCom's statutory proposal reaches Congress are the highest-leverage period for shaping the acquisition architecture. What gets designed into the command's concept of operations now will be operationally consequential for years.

Spartan X's work on edge AI architectures, multi-platform command integration, and AI verification has been built around exactly the questions AutoWarCom will need to resolve: validating autonomous system behavior under degraded conditions, establishing interoperable data standards across heterogeneous platforms, and defining T&E criteria that hold in joint environments. That experience is directly relevant to what Project Agincourt needs as it designs the command's organization, legislation and standards structure.