The Department of War's FY27 budget proposes $54.6 billion for the Defense Autonomous Warfare Group (DAWG). DAWG in its current form is an organizational construct rather than a unified command. It has consolidated programs that previously sat under Replicator and adjacent efforts, and it provides a budget line and a coordination function across the autonomous systems portfolio. What it does not yet have are the authorities of a unified command, the institutional independence of a separate service, or the force structure that would allow it to operate autonomous warfare capability rather than coordinate it. The institutional question worth working through is not what DAWG is today. It is what happens when DAWG evolves into or is replaced by an Autonomous Warfare Command (AWC) with real authorities, and what the services will do when that transition begins.
The current state matters because it is keeping the institutional fight artificially small. DAWG does not have the authorities that would force the services to give up programs, end strength, or mission ownership, which means the services have not yet had to mount the kind of resistance that defined the SOCOM and Space Force standups. The transition to AWC, whenever and however it happens, is the point at which that resistance will become serious. The analysis below is about what that fight will look like when it arrives.
For those with long memories, this dynamic has played out before. When Congress created Headquarters, US Special Operations Command in 1987, it did so over the active resistance of every service Chief. The services viewed SOCOM as a direct threat to their budgets, their end strength, and their institutional autonomy. That resistance lasted years, and in the Marine Corps' case delayed full participation until 2006, nearly two decades after SOCOM's activation. The creation of the Space Force followed a similar pattern, with the Air Force's most senior leaders publicly opposing the establishment of a new service until Congress overrode their objections through legislation.
An autonomous warfare command stood up with real authorities will face the same dynamics, except the mission space it claims is broader than either precedent. SOCOM pulled from a narrow slice of each service's most specialized units. The Space Force carved out a domain the Air Force had managed as a supporting function. AWC would touch every service's modernization priorities, acquisition programs, and vision of how each service fights. The institutional resistance will be proportional to the scope of what is being asked.
How New Commands Are Born
The creation of SOCOM is the most instructive precedent because the institutional dynamics are so thoroughly documented. Operation Eagle Claw in 1980, the failed hostage rescue attempt in Iran, killed eight service members and exposed a catastrophic inability to coordinate special operations across services. The Holloway Commission identified lack of unified command as the root cause. Even after that disaster, it took six more years of Congressional pressure to create SOCOM, because the services were unwilling to reform themselves.
Ultimately, it was Congress, not the Pentagon, that forced the creation. Jim Locher, the principal Senate Armed Services Committee staffer who helped draft both the Goldwater-Nichols Act and the Nunn-Cohen amendment, reported getting a great deal of pushback from a Pentagon still entrenched in its Cold War philosophy even after SOCOM was established. The final bill creating SOCOM was attached as a rider to the 1987 Defense Authorization Act, a move widely interpreted as Congress forcing the hand of the Department of Defense and the Reagan administration. Congress then had to pass two additional bills to ensure implementation because the Pentagon was slow-rolling compliance.
The resistance from the services was not abstract. Representative Daniel's original proposal would have created a national special operations agency headed by a civilian who would bypass the Joint Chiefs and report directly to the Secretary of Defense, specifically to keep the Joint Chiefs and the services out of the SOF budget process. That was too aggressive even for reformers, but it illustrates the depth of distrust that voluntary cooperation between services ran into. The compromise was the creation of Major Force Program 11, the SOF checkbook, which gave SOCOM control over its own resources and a degree of budget independence from the services. Without MFP-11, the services could have starved SOCOM through the normal budget process. Even with it, a 1991 RAND study, Providing Resources for Special Operations Forces, found that SOCOM had to negotiate literally hundreds of issues with the services and Defense agencies to guarantee essential support functions, and that disputes over whether expenses were service-common (paid by the services under MFP-2) or SOF-peculiar (paid by SOCOM under MFP-11) were frequent, contentious, and remained unresolved decades after SOCOM's creation.
The Marine Corps' resistance to SOCOM is the most vivid case study and the most relevant for understanding what AWC will face. The Corps consistently refused to contribute a dedicated component to SOCOM for nearly twenty years after SOCOM's founding. Marine leadership argued that Force Reconnaissance units already performed SOF-like missions while remaining organic to the Marine Air-Ground Task Force, and that detaching an elite Marine unit to a joint command would be to the detriment of the Marine Corps as a whole. Senior leaders feared that prized Marine units such as Force Recon would be prime targets for cherry picking if SOCOM were licensed to do so. Academic analysis from the Naval Postgraduate School described the Marine Corps' approach as an acknowledge and evade strategy to retain control of its Marines, prevent the creation of an elite within an elite in its ranks, and undermine Secretary Rumsfeld's eventual desire to create a Marine Corps special operations component. It took a direct order from Secretary Rumsfeld in the fall of 2005 to force the issue, with Commandant General James L. Jones brokering the eventual integration. MARSOC was officially activated on February 24, 2006, nineteen years after SOCOM's creation.
The Space Force followed a compressed but recognizable version of the same pattern. Vice Chairman of the Joint Chiefs General Paul Selva, Air Force Secretary Heather Wilson, and Air Force Chief of Staff General David Goldfein all publicly opposed the establishment of a new service. Congress and the President overrode that opposition through legislation.
The pattern is consistent across decades. Services protect three things above all else: end strength, budget share, and mission ownership. Any service-level change that threatens any one of these faces significant resistance. An autonomous warfare command stood up with real authorities threatens all three for every service simultaneously. Each service has spent the last several years building its own autonomous warfare infrastructure, with fully funded programs, dedicated organizations, acquisition pipelines, and institutional champions that will not be surrendered to a new joint command without a fight.
US Navy
The Navy arguably has the longest operational track record with autonomous systems of any service, with institutional depth across acquisition, experimentation, and force structure that predates the current autonomy push by several years.
The Navy's unmanned surface vessel effort began with separate Large Unmanned Surface Vessel (LUSV) and Medium Unmanned Surface Vessel (MUSV) programs, both of which drew sustained congressional scrutiny over propulsion reliability and the maturity of autonomous navigation. In 2025, the Navy merged LUSV and MUSV into the Modular Attack Surface Craft (MASC) program. MASC was itself cancelled in March 2026 and replaced with the MUSV Family of Systems (FoS) program, a marketplace-based acquisition approach under the Golden Fleet initiative seeking production-ready vessels rather than developmental prototypes. The rapid succession of program structures, LUSV and MUSV to MASC to MUSV FoS in under two years, demonstrates a service still working out how to acquire autonomous vessels at scale, but doing so with increasing urgency.
On the operational side, the Navy has been running autonomous systems in a real-world maritime environment since 2021, when 5th Fleet established Task Force 59 in Bahrain to integrate unmanned systems and AI into operations across the Arabian Gulf, Red Sea, Gulf of Oman, and parts of the Indian Ocean. TF 59 reached full operational capability in January 2023 and operated USVs from multiple vendors alongside partner nations through exercises like Digital Horizon and IMEX. In 2024, the task force transitioned into Task Group 59.1, shifting from experimentation to operational deployment. TG 59.1 deployed from Jordan into the Red Sea in a GPS-denied environment alongside a partner nation, demonstrating contested maritime operations that outside of Ukraine's endeavors in the Black Sea, most autonomous programs were still theorizing about.
The force structure buildout has been equally aggressive. USVRON-1 (Port Hueneme), USVRON-3 (Coronado), and USVRON-7 (stood up 2025) all sit under Surface Development Group ONE (SURFDEVGRU-1). USV Divisions 12 and 13 conduct unit-level training in San Diego. In early 2026, the Navy stood up three additional early command USV divisions and accepted delivery of upgraded prototypes Sea Hunter and Seahawk for integration with a carrier strike group. The Navy also demonstrated an SM-6 missile launch from the USV Ranger using a containerized VLS, proving that autonomous surface vessels can deliver lethal fires without a manned crew aboard.
In late 2025, the Navy established a Principal Acquisition Executive for Robotics and Autonomous Systems (PAE RAS) that could oversee up to 66 programs and approximately $19 billion in acquisitions over five years. Beyond surface vessels, the MQ-25 Stingray, MQ-4 Triton, and the XLUUV Orca represent autonomous programs developed independently across air, surface, and subsurface domains.
The PAE RAS reorganization is itself a statement of institutional intent. The Navy is consolidating its autonomous portfolio precisely so it cannot be disaggregated and handed to someone else. The service views maritime autonomy as integral to Distributed Maritime Operations and will argue that its unmanned fleet must remain under naval command to be effective in the maritime domain. The recent Saronic bid protest against NAVSEA's bundling of USV sustainment with LCS mission modules highlights a further complication. The Navy's own acquisition structures are still fighting over who controls USV programs internally, even before a joint command enters the picture.
United States Air Force
The Air Force has the deepest historical experience with armed unmanned systems of any service. The MQ-1 Predator entered service in 1995, the MQ-9 Reaper followed in 2007, and the supporting infrastructure of pilots, sensor operators, intelligence analysts, satellite networks, and ground control stations has been refined across nearly three decades of combat operations. That institutional base is the foundation the Air Force is now building on for Collaborative Combat Aircraft (CCA).
The Collaborative Combat Aircraft program is the centerpiece of the Air Force's current autonomous push. The Air Force launched the formal CCA program in 2023 and selected Anduril Industries and General Atomics in 2024 to build Increment 1 production-representative test articles. In March 2025, the service designated the prototypes the YFQ-42A (General Atomics, branded Dark Merlin) and the YFQ-44A (Anduril, branded Fury). GA-ASI's first pre-production YFQ-42A flew in August 2025. Anduril's YFQ-44A flew in October 2025. In December 2025, Northrop Grumman's Project Talon was designated the YFQ-48A and added to the Increment 1 competition.
The funding profile reflects how serious the Air Force is. FY25 provided $711.7 million in discretionary funding for CCA RDT&E plus $678 million in mandatory funding from reconciliation. FY26 added another $804.4 million combining discretionary and mandatory authorities. Total CCA investment is projected to approach $8.9 billion for Increment 1, currently scoped at 100 to 150 aircraft by the end of the decade. Increment 2 will be substantially larger, with total CCA force structure potentially reaching 1,000 aircraft or more.
The most consequential institutional development is the Air Force's separation of mission autonomy software from airframe hardware through the Autonomy Government Reference Architecture (A-GRA). A-GRA is a government-owned Modular Open System Approach designed explicitly to prevent vendor lock by decoupling the autonomy brain from the platform. Collins Aerospace provides the Sidekick autonomy software flying on the YFQ-42A. Shield AI provides Hivemind for the YFQ-44A. In February 2026, the Air Force validated A-GRA across both prototypes by demonstrating that mission autonomy software could be integrated across competing airframes from competing vendors. GA-ASI's YFQ-42A logged its first semi-autonomous airborne mission with Collins Sidekick on February 12, 2026, with a ground-based operator transmitting commands the aircraft executed for over four hours.
This software-hardware separation is the central ownership question for any joint command claiming authority over autonomous systems. The Air Force has structured A-GRA so the government owns the integrating architecture while industry competes on both airframes and autonomy stacks above and below it. The Air Force can swap algorithms across platforms, replace the autonomy stack on a CCA at software speed, and field, retire, or upgrade mission behaviors without redesigning the aircraft. The Air Force has positioned itself as the integrating authority over the entire CCA ecosystem. An autonomous warfare command claiming control of autonomy software would be claiming control of A-GRA itself, which the Air Force has built specifically to consolidate its own institutional ownership of the autonomy layer.
The control model is semi-autonomous, not fully autonomous. CCAs currently operate under the direction of a human pilot who functions as a mission commander, typically aboard an F-35 or eventually the F-47. Boeing was awarded the F-47 EMD contract in March 2025 (worth over $20 billion), with first flight planned for 2028 and IOC by 2029. The Air Force's vision is a manned-unmanned formation in which one pilot directs multiple autonomous wingmen, with the autonomy handling navigation, perception, dynamic planning, and weapon execution under human supervision. This is a deliberate human-on-the-loop architecture that preserves Air Force pilot authority while extending pilot reach.
The force structure to operate CCAs is being built in parallel. The Experimental Operations Unit at Nellis AFB was activated as a fully operational squadron-equivalent in June 2025 under Eglin's 53rd Wing. Beale AFB was selected in May 2025 as the preferred location for a CCA Aircraft Readiness Unit. A production decision for Increment 1 is expected in the current fiscal year.
The Air Force will resist any joint command claiming authority over CCA, A-GRA, or its autonomy software stack, and the resistance will be more institutionally entrenched than the Navy's for three reasons. CCA is inseparable from the manned fighter fleet by design; removing it from Air Force control would require rebuilding the manned-unmanned teaming concept from scratch. The autonomy software stack is becoming Air Force institutional property through A-GRA, and ceding it would surrender precisely what the Air Force has been working to capture. CCA is the Air Force's bid to remain the dominant air superiority service in an era of shrinking manned fighter procurement. The historical analog is U.S. Cyber Command, which the Air Force resisted on similar grounds before settling for a service-component model. An autonomous warfare command would face the same negotiation, with the Air Force arriving at it holding the most institutionally consolidated autonomy program in the Department.
US Army
The Army has the broadest unmanned portfolio in the Department of War, spanning small UAS at the squad level, medium UAS at the brigade level, unmanned ground vehicles at multiple weight classes, launched effects across air and ground, and loitering munitions at every echelon. No other service has fielded autonomy across ground and air domains at this scale, and no other service has pushed unmanned capability this deep into small unit tactics. For an autonomous warfare command, the Army is both the largest prize and the hardest to consolidate, because Army autonomy is not concentrated in a single program but instead dispersed into every brigade combat team in the force.
The unmanned ground vehicle portfolio is the most distinctive Army capability and has no analog in the Navy or Air Force. The Robotic Combat Vehicle (RCV) program is the centerpiece. The Army awarded Textron Systems's Ripsaw 3 the Phase II contract in March 2025, with full-system prototypes scheduled for FY26, a production decision in FY27, and first unit fielding in FY28. RCV is being developed in light, medium, and heavy variants to serve as scouts, escorts, and fires platforms. Alongside RCV, General Dynamics Land Systems produces the Small Multipurpose Equipment Transport (SMET) robotic mule, with a Medium MMET competition expected in 2026. An Autonomous Multi-domain Launcher (AML) and a competing Raytheon-led DeepStrike platform are being experimented with as autonomous launchers for long-range precision fires.
The RCV autonomy story is where platform-versus-software ownership becomes contested in a way that parallels the Air Force's A-GRA debate but is slightly different. While the Army has been developing an internal autonomy stack called the Robotic Technology Kernel (RTK) for years, industry raised persistent concerns that RTK is not mature enough for operational UGVs and that the Army is too institutionally committed to an in-house solution. In response, DIU launched a RCV Software Acquisition Pathway in 2024, selecting eight industry partners to prototype commercial autonomy: Overland AI, Forterra, Neya Systems, and Kodiak Solutions on navigation; Scale AI and Applied Intuition on machine learning; and Palantir and Anduril as software system integrators. The Army is now running RTK and the commercial SWP pathway in parallel, which means the service has not yet decided whether autonomy ownership lives in government hands or industry hands. That undecided question is precisely the kind of ownership ambiguity an autonomous warfare command could attempt to resolve, and it is also precisely the kind of question the Army has every incentive to resolve internally first.
On the aerial side, the Army has organized its small UAS portfolio by echelon in a way that no other service has attempted. At the squad level, the Soldier-Borne Sensor program has fielded the Teledyne FLIR Black Hornet since FY19. At the platoon level, the Purpose Built Attritable System program is fielding FPV drones patterned on Ukrainian usage, with the Drone Dominance Program's Gauntlet seeking 300,000 progressively cheaper munitions by 2028. At the company level, the Tranche 1 small UAS selection includes the Anduril Ghost-X and the Precision Drone Works C-100. At the battalion and brigade level, Short Range Reconnaissance UAS (Skydio X10D) are being fielded through Transformation in Contact brigades. Above that, FTUAS is replacing the RQ-7 Shadow, and the MQ-1C Gray Eagle continues to provide medium-altitude capability through Combat Aviation Brigades. Loitering munitions sit alongside this portfolio. LASSO equips infantry brigades with Switchblade 600s, and OPF-L pursues mass-producible squad and platoon munitions.
The operational concept tying all of this together is Transforming in Contact (TiC) and Human-Machine Integrated Formations (HMIF). TiC was launched in 2024 as a deliberate experimentation framework. The 2nd Brigade, 101st Airborne Division and the 3rd Brigade, 10th Mountain Division were the initial TiC brigades. The 1st ABCT, 3rd Infantry Division employs a Multi-Effects Company patterned on multi-functional reconnaissance teams: three hunter-killer platoons focused on drones, electronic warfare, and robotics. HMIF is the operational endpoint, envisioned as the first integrated manned-unmanned formations in 2027, including a common controller for multiple robotic systems.
Two external drivers are accelerating all of this. Secretary Pete Hegseth's April 30, 2025 memorandum directed the Army to field unmanned systems and ground and air launched effects in every Division by the end of 2026. Replicator 1 and Replicator 2, now consolidated under DAWG following the August 2025 Hegseth memo, have pushed autonomous systems into Army hands at scale. An April 2026 Small Wars Journal essay, Transforming in Contact: The Army Needs an Unmanned Systems Command Now, reflects active internal Army discussion about creating a dedicated Army unmanned systems command, which would solidify service ownership of the portfolio before any joint consolidation could take place.
The Army will resist any joint autonomous command claiming authority over its portfolio, but the resistance will play out differently than the Navy or Air Force will play it out. Unmanned systems are distributed across every echelon from squad through division, woven into the brigade combat team structure. Pulling unmanned out of the BCT would require restructuring the entire BCT. The Army is actively pursuing its own internal unmanned systems command, and if that command stands up before a joint autonomous warfare command consolidates, it will become a competing institutional claim on the same portfolio. The Army's UGV portfolio is unique. No other service has the depth of ground robotics investment, and an autonomous warfare command would struggle to absorb RCV, SMET, MMET, and AML without absorbing the Army ground combat ecosystem that surrounds them.
The practical implication is that the Army may be the service most willing to cooperate on air-domain autonomy consolidation (where it overlaps with Air Force and Marine Corps efforts) and the service most willing to fight on ground-domain autonomy consolidation (where it has unique equities). An honest implementation plan for AWC will have to anticipate that asymmetry and probably concede UGV authority to the Army while negotiating access to Army autonomy software, training infrastructure, and operational experimentation venues.
United States Marine Corps
Of all the services, the Marine Corps has built the most institutionally entangled relationship with unmanned systems. For the Navy and Air Force, autonomous platforms are major programs nested inside larger force structures. For the Marines, unmanned systems are the load-bearing element of the entire post-2020 force design. Removing them, or ceding them to a joint command, would not just reduce a portfolio. It would dismantle the operating concept the service has spent six years rebuilding itself around.
Force Design (originally Force Design 2030, now simply Force Design) was introduced by General David Berger in March 2020 as a deliberate restructuring of the Marine Corps for distributed maritime operations against China. The plan divested the entire tank fleet, cut tube artillery in favor of rocket artillery, and reshaped infantry battalions around small distributed units operating inside an adversary's weapons engagement zone. The replacement capabilities are almost entirely unmanned: long-range loitering munitions, organic precision fires, small UAS at the squad level, NMESIS anti-ship missile launchers, and a family of unmanned aerial systems integrated at every echelon. The Marine infantry squad was restored to 13 Marines specifically to include a drone operator. Infantry battalions added a Fires and Reconnaissance Company built around small UAS and rockets. The Marine Littoral Regiment construct, which converted the 3rd, 4th, and 12th regiments into multi-domain formations supporting the MEF, is structurally dependent on unmanned reconnaissance, surveillance, anti-ship targeting, and electronic warfare. 3rd MLR achieved FOC in Hawaii in 2023, 12th MLR is reaching FOC in Okinawa in 2026, and the MLR is now the model for how the Marine Corps fights in the Pacific.
The Marine Corps Attack Drone Team (MCADT), established January 3, 2025 by the Commanding Generals of Training Command and the Marine Corps Warfighting Laboratory, is the most visible recent addition (MARADMIN 235/25). Based at Weapons Training Battalion at Quantico, MCADT was created as a direct response to the proliferation of armed FPV drones in Ukraine. The unit conducted the Marine Corps' first kinetic live-fire FPV drone strike at Quantico on April 24, 2025, at a cost of less than $5,000 per strike compared to over $80,000 for current infantry battalion organic missiles, with effective ranges out to 20 kilometers compared to 1 to 2 kilometers for current squad-level systems. MCADT partnered with the Marine Corps Tactics and Operations Group to draft the service's first Armed Drone Employment TTPs pamphlet. It established the Service's first Armed FPV Drone Employment Competitions, held the inaugural competition at Quantico in October 2025, and ran follow-on competitions at Camp Hansen, Okinawa in December 2025 and MCB Hawaii in January 2026. MCADT trained and certified Marines from 22nd MEU(SOC) on FPV drones during a November 2025 exercise at Camp Santiago, Puerto Rico, marking the first time forward-deployed Marines were certified on FPV drones as a unit capability rather than as an experimental skill.
The institutional placement matters. MCADT sits under Weapons Training Battalion as a tenant unit alongside the Marine Corps Shooting Team, which has 124 years of marksmanship lineage. This is a deliberate signal that the Marine Corps treats armed FPV drones as an extension of individual weapons proficiency rather than as a separate aviation discipline. That placement creates an institutional model that is very different from how the Navy or Air Force has organized unmanned forces. Marine FPV expertise lives with the rifleman, not in a standalone unmanned squadron. A joint autonomous warfare command claiming authority over small UAS would be claiming authority over what the Marine Corps now considers an organic infantry weapon.
At the high end of the autonomy stack, the Marine Corps has been working the Kratos XQ-58 Valkyrie longer than the Air Force has. The Marine Corps ordered the first two XQ-58s in December 2022 under the Penetrating Affordable Autonomous Collaborative Killer Portfolio (PAACK-P) program and conducted first test flights at Eglin in October 2023. The service paired the Valkyrie with F-35Bs from VMFA-214 for electronic attack and SEAD roles. In January 2026, the Marines made the XQ-58 a program of record under the MUX TACAIR program, awarding a $231.5 million prototype contract to a Northrop Grumman and Kratos team. Northrop Grumman is the prime contractor and is integrating its Prism open-architecture autonomy software (the same Prism stack flying on Northrop's YFQ-48A in the Air Force's CCA Increment 1 competition) onto the Kratos airframe. The Marine Corps formally designated the XQ-58A as a Collaborative Combat Aircraft, the same designation the Air Force uses.
That parallel is significant for how an autonomous warfare command would have to handle ownership questions. The Marine Corps and the Air Force are running CCA programs with overlapping vendors, overlapping autonomy software, overlapping platforms, and overlapping operational concepts. Today the two services are de-conflicting through coordination. A joint command attempting to consolidate either the airframe acquisition or the autonomy software stack would have to decide whether the Marine Corps version of CCA is its own program or a subset of the Air Force program, and that decision would force a fight over which service controls the autonomy that flies on a Marine F-35B's wingman.
Beyond CCA and MCADT, MARSOC operates as the Marine component of SOCOM, with its own organic small UAS, ground robotic systems, and autonomous mission packages developed under SOF acquisition authorities. The MARSOC integration story described earlier is the most cautionary precedent. Nineteen years of resistance, an acknowledge and evade posture, and integration only under direct SECDEF order is not an outlier. It is the Marine Corps' baseline response to any external command claiming authority over Marine forces. There is no reason to expect a different posture if a similar consolidation is attempted on autonomous systems, and several reasons to expect it will be more intense because Force Design has embedded unmanned systems far more deeply into the Marine Corps' identity than special operations ever were.
The Marine Corps will resist any joint autonomous command for three structural reasons distinct from the Navy and Air Force objections. The MAGTF is doctrinally and culturally sacred. Unmanned systems are now distributed across all four MAGTF elements (command, ground, aviation, logistics). Pulling unmanned out of the MAGTF would not be a clean acquisition transfer; it would be an attack on the operating concept. Force Design is the Commandant's signature institutional initiative, and the unmanned portfolio is the centerpiece of Force Design. Ceding that portfolio to a joint command would politically humiliate the service in a way that the Navy ceding USV programs or the Air Force ceding CCA would not, because for the Marines the unmanned portfolio is the future of the service rather than one program line among many. The Marine Corps treats unmanned systems as small unit enablers rather than as a standalone capability category. The institutional placement of MCADT under Weapons Training Battalion is the clearest expression of this. Armed drones are individual weapons in the Marine view, and individual weapons stay with the Marines who carry them.
The MARSOC precedent and the institutional placement of MCADT together suggest the Marine Corps would respond to a proposed autonomous warfare command the same way it responded to SOCOM. The service would acknowledge the requirement publicly, contribute the minimum necessary participation to avoid being overruled, and slow-roll the consolidation through a combination of doctrinal arguments about the MAGTF, structural arguments about Force Design coherence, and cultural arguments about the inseparability of the rifleman from the rifleman's tools. If the recent MARSOC history is any guide, that resistance could last fifteen to twenty years before the Marine Corps voluntarily integrates.
What the Pattern Tells Us About the Transition Ahead
The $54.6 billion in DAWG funding tells us the Pentagon is moving toward an autonomous warfare command. DAWG in its current form is a transitional structure. It coordinates programs, consolidates Replicator-era investments, and provides a budget line, but it does not yet have the authorities of a unified command. The institutional fight that defined SOCOM and the Space Force has not begun in earnest because DAWG cannot yet demand from the services what AWC will eventually demand. That fight begins when the transition to a real autonomous warfare command starts.
Every historical precedent says the transition will be long, institutional, and decided as much by Congressional action as by Pentagon planning. SOCOM took a catastrophic failure that killed eight service members and six years of legislative pressure before Congress forced its creation over service resistance. Even then, the Marine Corps held out for another nineteen years. The Space Force took direct Presidential and Congressional action over the public opposition of the Air Force's most senior leaders. Both organizations are now vital to national security, and both would not exist if the services had been left to reform themselves.
The transition from DAWG to a real AWC does not yet have either forcing function. There is no catastrophic failure that makes the status quo indefensible. There is not yet a Congressional mandate that overrides service resistance. What there is is a budget number, a growing threat in the Pacific, an accelerating technological trajectory that Ukraine has validated in combat, and the institutional memory of what happens every time the Pentagon is left to manage its own organizational transformation.
Every service will argue that its existing autonomous programs are working. Every service will point to MASC, CCA, JIATF-401, RCV, or MUX TACAIR as evidence that autonomous warfare is already being addressed within existing structures. Those arguments are not wrong. The programs are real and many of them are producing results. The question is whether independent, service-centric development produces a joint autonomous force that can fight as a coherent whole, or whether it produces four separate autonomous enterprises that work brilliantly within their own domains and fail at the seams between them.
That question has been answered before. The answer is why SOCOM exists, why the Space Force exists, and why autonomous warfare will eventually get its own command. The variables are how long the transition takes, how much institutional friction is burned in the process, and whether the forcing function is a strategic failure that could have been prevented or a Congressional mandate pursued before the next fight begins.