The memorandum is dated 9 January 2026 on the letterhead of the Secretary of War and addressed to senior Pentagon leadership, the commanders of the combatant commands, and defense agency and field activity directors. It was posted to the Department's media server on 12 January 2026. Its subject line reads "Artificial Intelligence Strategy for the Department of War"; its opening heading is "Accelerating America's Military AI Dominance." It is the originating instrument of the contract dispute recorded at Anthropic v. United States (Pentagon ban challenge), and the memorandum CNN identified in September 2026 as the source of the department-wide model-access directive. For the Department's AI programs generally, see DOD — Department of Defense (AI Deployer); for the strategy as a policy instrument, Artificial Intelligence Strategy for the Department of War (2026).
Framing
The memorandum opens by quoting Executive Order 14179: "It is the policy of the United States to sustain and enhance America's global Artificial Intelligence (AI) dominance in order to promote human flourishing, economic competitiveness, and national security." It asserts that "AI-enabled warfare and AI-enabled capability development will re-define the character of military affairs over the next decade," calls the transformation "a race-fueled by the accelerating pace of commercial AI innovation coming out of America's private sector," and directs the Department, "aligned with America's AI Action Plan" (America's AI Action Plan), "to accelerate America's Military AI Dominance by becoming an 'AI-first' warfighting force across all components, from front to back."
Four means are named: unleashing experimentation with leading US models department-wide and rewarding "AI-first re-conceptions of legacy approaches"; identifying and eliminating bureaucratic barriers "which are vestiges of legacy information technology and modes of warfare"; concentrating investment on what the memorandum calls America's "core asymmetric advantages" in compute, model innovation, entrepreneurial dynamism, capital markets and "combat-proven operational data from two decades of military and intelligence operations that no other military can replicate"; and executing a set of Pace-Setting Projects.
The seven Pace-Setting Projects
The projects are to be run in fiscal year 2026 under "single accountable leaders, aggressive timelines, measurable outcomes, and rapid iteration where failure accelerates learning and improvement," each led by a program leader with a sponsoring organization, demonstrated monthly to the Deputy Secretary of War and the Under Secretary of War for Research and Engineering, with initial demonstration by transition-partner users within six months of the memorandum.
| # | Project | Mission area | Stated purpose |
|---|---|---|---|
| 1 | Swarm Forge | Warfighting | Competitive mechanism to discover, test and scale ways of fighting with and against AI-enabled capabilities, pairing elite warfighting units with technology innovators |
| 2 | Agent Network | Warfighting | AI agent development for battle management and decision support, "from campaign planning to kill chain execution" |
| 3 | Ender's Foundry | Warfighting | AI-enabled simulation and sim-dev / sim-ops feedback loops |
| 4 | Open Arsenal | Intelligence | Accelerating the technical-intelligence-to-capability pipeline, "turning intel into weapons in hours not years" |
| 5 | Project Grant | Intelligence | Transforming deterrence "from static postures and speculation to dynamic pressure with interpretable results" |
| 6 | GenAI.mil | Enterprise | "Democratizing AI experimentation and transformation across the Department by putting America's world-leading AI models directly in the hands of our three million civilian and military personnel, at all classification levels" |
| 7 | Enterprise Agents | Enterprise | A playbook for rapid, secure agent development and deployment in enterprise workflows |
Each military department, combatant command, defense agency and field activity is directed to identify within 30 days at least three projects to "fast-follow" the seven. The Chief Digital and AI Office is to track and rank the extended set "by speed and impact," reporting monthly. Classified annexes provided separately are to be accelerated in parallel.
Enablers and deadlines
Funding. The memorandum cites program funding across the services, resources "provided by Congress in the form of One Big Beautiful Bill," and expanded budget-withhold flexibility through a Joint Acceleration Reserve, and says allied and partner access, capabilities and investments will be leveraged consistent with the AI Action Plan's international diplomacy line.
Compute. The Department is to invest in expanding access to compute "from datacenters to the edge," leveraging "the hundreds of billions in private sector capital investment being made in America's AI sector through our growing array of creative partnerships," to work with interagency partners on technical standards for new secure datacenters, and to support and use the American Science and Security Platform being built under the Genesis Mission.
Data. The CDAO is directed to enforce the "DoD Data Decrees." Military departments and components are to maintain federated data catalogs exposing system interfaces, data assets and access mechanisms at all classification levels under the May 2021 "Creating Data Advantage" memorandum, and to deliver current catalogs within 30 days. The Under Secretary of War for Intelligence and Security is to give intelligence data parallel treatment on the same timeline. The CDAO "is authorized to direct release of any DoW data to cleared users with valid purpose, consistent with security guidelines," and denials of CDAO data requests must be justified to the Under Secretary for Research and Engineering within seven days, who will remediate or escalate. The memorandum's stated rationale: "Our data advantage is meaningless if our developers and operators cannot exploit it."
Talent. Special hiring and pay authorities are directed department-wide, with each component to submit AI hiring and talent-development plans to the Under Secretary of War for Personnel and Readiness within 60 days, for approval, denial or modification within 30 days after.
Acceleration expectations
Six expectations are set out under the heading "Acceleration Expectations," each carrying a directive.
- Speed Wins. "We must accept that the risks of not moving fast enough outweigh the risks of imperfect alignment." The CDAO is to establish deployment-velocity and operational cycle-time metrics for all Pace-Setting Projects.
- AI Model Parity. "The Department cannot be working off models that are months or years old." The CDAO is to establish a delivery and integration cadence with vendors "that enables the latest models to be deployed within 30 days of public release," which "shall be a primary procurement criterion for future model acquisition."
- Wartime Approach to Blockers. Blockers to data sharing, Authorizations to Operate, test and evaluation, certification, contracting and hiring are to be eliminated, treating "risk tradeoffs, 'equities', and other subjective questions as if we were at war." The Under Secretary for Research and Engineering is to establish a monthly "Barrier Removal Board" with authority to waive non-statutory requirements.
- Competition > Centralized Planning. "Small, accountable teams will win over process." The CDAO is to establish AI system usage and mission-impact metrics, and "decisions about future resourcing and deprecation of associated capabilities will principally be made on the basis of these metrics."
- AI-Native Warfighting. Each service chief and combatant commander is to designate an AI Integration Lead within 30 days; the Joint Staff is to designate a senior official to monitor service AI warfighting concept development and report quarterly. "Exercises and experiments that do not meaningfully incorporate AI and autonomous capabilities will be reviewed by the Director of Cost Assessment and Program Evaluation for resourcing adjustment."
- Modular Open Architectures. Program managers acquiring AI capabilities are to enforce Modular Open System Architectures and the Data Decrees, "exposing modular interfaces and associated documentation sufficient for third-party integration without prime contractor support."
The "Responsible AI" subsection
Under the heading "Clarifying 'Responsible AI' at the DoW — Out with Utopian Idealism, In with Hard-Nosed Realism," the memorandum states that "Diversity, Equity, and Inclusion and social ideology have no place in the DoW, so we must not employ AI models which incorporate ideological 'tuning' that interferes with their ability to provide objectively truthful responses to user prompts," and that "the Department must also utilize models free from usage policy constraints that may limit lawful military applications." It issues three directives: the CDAO is to establish "benchmarks for model objectivity as a primary procurement criterion" within 90 days; the Under Secretary of War for Acquisition and Sustainment is to "incorporate standard 'any lawful use' language into any DoW contract through which AI services are procured" within 180 days; and the CDAO is to align all existing departmental AI policy guidance with the memorandum.
This is the provision that the restrictions in Anthropic's usage policy ran against, and the passage the summary-judgment opinion in Anthropic v. United States (Pentagon ban challenge) quotes (Anthropic PBC v. U.S. Department of War — Order on Cross Motions for Summary Judgment (N.D. Cal., August 27, 2026)).
Dating
Accounts of the memorandum have given two dates. The Verge's report of 26 May 2026 described the originating instrument as a memo of 12 January 2026 demanding renegotiation of AI contracts on "any lawful use" terms and arguing that "the risks of not moving fast enough outweigh the risks of imperfect alignment." Both passages appear in this memorandum, which is dated 9 January 2026 on its face and posted at a URL path dated 12 January — the same date the opinion in the Anthropic litigation assigns to the strategy memorandum's substance while dating the document itself to 9 January. On the face of the primary text, the memorandum is one instrument dated 9 January 2026 and published on 12 January.
Provenance
Captured on 19 September 2026 from media.defense.gov, the Department's official media server, through the Firecrawl CLI after a direct fetch returned HTTP 403 to a non-browser client. CNN's report of 18 September 2026 links this exact URL as the memorandum announcing the AI acceleration strategy and quotes the GenAI.mil passage, which appears in the text. The capture covers the letterhead, date, addressees, subject and all sections through "Becoming An AI-First Department"; the signature block did not extract, and the text layer carries OCR-style artifacts ("Do W" for "DoW", "Al" for "AI", "st:curity"), silently corrected in the quotations above where the intended reading is unambiguous. The classified annexes referenced in the memorandum are not public.
Relationships
- supports: Artificial Intelligence Strategy for the Department of War (2026) — the policy instrument this text establishes.
- supports: Anthropic v. United States (Pentagon ban challenge) — supplies the "any lawful use" directive at issue.
- depends-on: Executive Order 14179 — Removing Barriers to American Leadership in Artificial Intelligence, America's AI Action Plan — the policy authorities the memorandum invokes.
- related: Pete Hegseth — the signatory.
- deploys-in: DOD — Department of Defense (AI Deployer) — the department the memorandum binds.
- related: DoD Directive 3000.09 — Autonomy in Weapon Systems — the standing autonomy-in-weapons directive the memorandum does not amend.
- related: Autonomous Weapons — the Agent Network project's "kill chain execution" language.
- contradicts: the account dating the originating memorandum to 12 January 2026 (Source: theverge.com).