Investigate how major U.S. federal agencies…

Major U.S. federal agencies have responded to 2025 directives—primarily Executive Order 14179 (Removing Barriers to American Leadership in Artificial Intelligence, January 2025), OMB Memorandum M-25-21 (Accelerating Federal Use of AI through Innovation, Governance, and Public Trust, April 2025), and the July 2025 America’s AI Action Plan—by issuing agency-specific AI strategies or compliance plans.[1][2]

These documents emphasize accelerating responsible AI adoption, removing bureaucratic barriers, strengthening governance and public trust, and aligning with national goals of U.S. AI dominance. Most “covered agencies” (per 44 U.S.C. § 3502) published initial strategies by the September 30, 2025 deadline, with follow-on detailed policies due by December 29, 2025.[1]

DoD’s January 2026 AI Acceleration Strategy: The Most Recent and Aggressive Update

The Department of Defense (DoD, referred to as the Department of War in some 2026 documents) released its most recent major update on January 9–12, 2026: the Artificial Intelligence Acceleration Strategy (a short memorandum directing an “AI-first” warfighting force). This builds on prior efforts (e.g., the 2023 Data, Analytics, and AI Adoption Strategy) but shifts toward wartime-speed execution aligned with the national AI Action Plan.[3][4]

It prioritizes rapid integration of commercial frontier models, elimination of legacy barriers, and leveraging U.S. asymmetric advantages (compute, data from operations, capital markets, and talent). Key mechanisms include seven “Pace-Setting Projects” (PSPs) spanning warfighting (e.g., Swarm Forge for AI swarms, Agent Network for battle management), intelligence (e.g., Open Arsenal for faster intel-to-weapon pipelines), and enterprise operations (e.g., GenAI.mil for department-wide access and Enterprise Agents).[5]

• Core directives: Deploy latest models within 30 days of public release (“AI Model Parity”); enforce data access via “DoD Data Decrees” with rapid CDAO authority; expand compute via public-private partnerships; accelerate talent hiring; create a monthly Barrier Removal Board; measure success via deployment velocity, usage metrics, and mission impact.[3]
• Implications for competitors/contractors: Emphasis on speed favors agile vendors offering rapid updates, classified-environment compatibility, and integration support. Bureaucratic relief creates opportunities but requires alignment with PSPs and metrics.

September 2025 Wave: DHS, HHS, GSA, and Other Agencies’ Responses to M-25-21

Dozens of agencies released strategies or compliance plans around the September 30, 2025 deadline. Notable examples include DHS, HHS (encompassing NIH), GSA, DOE, State, VA, CFPB, and NARA.[1]

DHS AI Strategy (September 2025) focuses on a three-year maturity roadmap with mission-specific use cases and three maturity goals.[6][6]

• Use cases: Border security (autonomous surveillance, risk assessment); customs/travel (facial recognition/biometrics); transnational crime investigations (image recognition, translation); cybersecurity (anomaly detection via CyberSentry and CISA tools); passenger security (TSA item detection, staffing optimization); immigration (screening, E-Verify, translation).[6]
• Maturity goals: (1) AI-enabling infrastructure (consolidated IT, AI-as-a-Service APIs, testing/evaluation, continuous monitoring); (2) Data governance (quality, traceability, reuse/sharing); (3) Strategic mission delivery (AI-ready workforce, R&D, governance/risk management, resource tracking).[6]

HHS AI Strategy (issued September 30, 2025; publicized December 2025) adopts a “OneHHS” approach across divisions (CDC, CMS, FDA, NIH, etc.) with five pillars.[7][8]

• Pillars: (1) Governance and risk management for public trust (roles, inventories, transparency, privacy); (2) Infrastructure and platforms (OneHHS AI integrated Commons for shared data/compute/models/testbeds); (3) Workforce development and burden reduction (skills, tools to cut admin work); (4) Health research and reproducibility (Gold Standard Science principles); (5) Care and public health delivery modernization (outcomes-focused integration).[7]
• NIH-specific elements remain integrated (e.g., via the broader HHS plan and the separate 2025–2030 NIH Strategic Plan for Data Science, which includes AI goals); a standalone NIH AI Strategic Plan was still under development following a June 2025 RFI.[9]

GSA published its Strategies for OMB Memorandum M-25-21 on September 30, 2025, covering use cases, governance (Chief AI Officer, boards, oversight), training/community-building (e.g., demo days), and a compliance plan. GSA also supports government-wide efforts like an AI Center of Excellence and procurement guidance.[10][10]

Other agencies (VA, DOE, etc.) issued similar compliance plans emphasizing use-case inventories, risk frameworks, and alignment with M-25-21 pillars.[1]

Top Cross-Agency Themes

Agency strategies converge on several mechanisms and priorities, reflecting the national shift from caution to accelerated, mission-aligned adoption:

• Barrier removal and speed of adoption — Common emphasis on experimentation, scaling pilots to enterprise use, and streamlining procurement/IT processes (most explicit in DoD’s “wartime approach” and model parity rules).[3]
• Governance, risk management, and trust — Mandatory or recommended AI inventories, risk rubrics/assessments, transparency, privacy/civil rights protections, and Chief AI Officer roles or boards.
• Infrastructure and data foundations — Shared platforms (e.g., AI-as-a-Service, Commons), compute access/expansion, data quality/FAIR principles, traceability, and reuse/sharing.
• Workforce and talent — Broad AI literacy training plus targeted recruitment for data science, ML engineering, and ethics roles; reducing administrative burdens via AI tools.
• Mission-tailored applications with measurable outcomes — Defense (lethality, decision superiority), health (research translation, care delivery, “Make America Healthy Again”), homeland security (border, cyber, investigations); metrics for progress and impact.
• Alignment with national AI dominance goals — References to the AI Action Plan, EO 14179, and public-private partnerships; reduced focus on prior-era risk emphases in favor of innovation.

Implications for New Entrants or Competitors

Agencies are building internal capacity (e.g., shared infrastructure, training) while relying on contractors for specialized models, integration, and use-case delivery. Success requires demonstrating rapid deployment, compliance with agency-specific governance/risk frameworks, support for data/compute ecosystems, and measurable mission outcomes. DoD’s emphasis on 30-day model updates and PSP participation, combined with government-wide procurement streamlining (M-25-22), rewards vendors who can operate at commercial speeds within federal constraints. Ongoing updates (e.g., annual revisits through 2036 per some guidance) mean strategies will evolve; monitoring Chief AI Officers and public inventories will be key.[11]

Overall, the federal approach has shifted decisively toward operationalizing AI as a core capability, with September 2025 marking a broad baseline and DoD’s January 2026 strategy representing the leading edge of acceleration.

---

Recent Findings Supplement (June 2026)

DoD (rebranded references as Department of War) released its most recent major agency-level AI strategy in January 2026.[1][2]

This "Artificial Intelligence Strategy for the Department of War" (memorandum dated January 9, 2026; publicly launched January 12) directs an "AI-first" warfighting force to achieve "Military AI Dominance." It responds to the national America's AI Action Plan and related executive orders by emphasizing rapid experimentation, barrier removal, and leveraging U.S. private-sector advantages (compute, models, capital, and unique combat data). The mechanism centers on seven initial "Pace-Setting Projects" (PSPs) in warfighting, intelligence, and enterprise areas—each with single accountable leaders, aggressive timelines (initial demos within six months), monthly progress reporting to senior leadership, and real-time sharing of foundational enablers (infrastructure, data, models, policies, talent) across the Department. Priorities include "speed wins" (cycle-time metrics), AI model parity (latest models deployed within 30 days of public release), a "wartime approach" to blockers (e.g., data sharing, ATOs via a Barrier Removal Board), internal competition via small teams and metrics-driven resourcing, and AI-native warfighting concepts. It re-focuses the Chief Digital and AI Office (CDAO) for FY2026 execution, enforces data catalogs and access, and directs talent plans.[1]

This is the most recent comprehensive strategy release among major agencies surveyed. For entrants or competitors, it signals DoD's shift to accelerated, metrics-driven adoption with heavy emphasis on speed and private-sector integration, creating opportunities in PSP-aligned areas (e.g., AI agents, simulations, tech-to-capability pipelines) but requiring alignment with rapid-update and security standards.

HHS released its department-wide AI Strategy in late 2025 (with implementation details and related NIH work extending into 2026), emphasizing a "OneHHS" collaborative infrastructure model.[3][4]

The strategy (issued around September 30, 2025, with December announcements and 2026 updates) focuses on internal operations, efficiency, research, and public health while fulfilling OMB directives. It establishes governance and risk management, inventories of AI use cases, a shared "OneHHS AI integrated Commons" for data/resources/compute/models/testbeds across divisions (CDC, CMS, FDA, NIH, etc.), workforce upskilling to reduce burdens, embedding "Gold Standard Science" for research translation/reproducibility, and outcomes-focused modernization of care/public health.[3]

NIH-specific developments include the 2025–2030 Strategic Plan for Data Science (with explicit AI goals around software/methods and infrastructure) and March 26, 2026, insights from AI Assurance Lab pilots. These pilots identified gaps in standardized guidance, benchmarks, testing/evaluation methods, and playbooks, leading to a recommendation for a collaborative NIH AI Assurance Lab (potentially CORAL) to generate real-world, iterative resources via partnerships (e.g., with MITRE) and address resource-intensive custom workflows.[5][6]

Implications: Agencies like HHS/NIH prioritize shared infrastructure and assurance to scale responsibly in biomedical contexts. Competitors can engage via the Commons or assurance initiatives but must navigate privacy, ethics, and reproducibility requirements.

GSA published updated AI resources and 2025 use-case inventories in May 2026 as part of ongoing OMB M-25-21/M-25-22 compliance.[7]

GSA's May 11, 2026, resources page details compliance strategies and plans for the OMB memoranda on accelerating federal AI use (innovation, governance, public trust) and efficient AI acquisition. It includes a curated 2025 GSA AI use cases inventory (covering applications like document review, risk management, and sales processes), AI governance structures for risk mitigation, and an AI Community of Practice for cross-agency collaboration and workforce capacity. Earlier 2025 actions (e.g., USAi platform launch) supported the national AI Action Plan, with continued focus on FedRAMP authorizations for AI tools.[7]

This positions GSA as a facilitator for government-wide adoption. For others, it highlights opportunities in compliant acquisition, shared platforms, and use-case sharing, with emphasis on governance documentation.

DHS (September 2025 strategy with December 2025 page updates), VA, Treasury, and others issued or updated compliance plans and inventories aligned to OMB M-25-21, with activity into early 2026.[8][9]

DHS's AI Strategy (September 2025) outlines a three-year approach to barrier removal, AI maturity enhancement, transparency/accountability, and scaling use cases per OMB guidance; related compliance plans address high-impact AI risk management (deadlines into 2026).[10] VA maintains an AI Strategy and compliance plan with 2025 inventory processes and April 2026 high-impact deadlines. Treasury released January 2026 AI use-case inventories and launched an AI Innovation Series in March 2026 focused on strategy/governance in financial contexts.[11]

These reflect a wave of agency responses to the same federal directive.

Top cross-agency themes from recent plans emphasize alignment with national directives, practical acceleration mechanisms, and responsible scaling. Common priorities include: robust governance/risk management and public AI use-case inventories (with high-impact focus and deadlines like April 2026); infrastructure investments (shared commons, compute access, data catalogs); workforce/talent development; responsible/secure/trustworthy AI (ethics, privacy, assurance standards); mission-specific acceleration (e.g., warfighting speed for DoD, health outcomes for HHS); and competition/partnerships with industry. Many explicitly tie to OMB M-25-21/M-25-22 and the America's AI Action Plan, shifting from high-level vision to implementation roadmaps, compliance plans, and measurable projects.[12]

A June 2026 executive order further directs interagency actions (e.g., involving DHS/CISA on AI security).[13]

For competitors or new entrants, success requires demonstrating alignment with these themes—particularly rapid, governed deployment, data/AI infrastructure compatibility, and measurable mission impact—while navigating agency-specific compliance and deadlines. Recent activity shows a maturation from strategy to execution, with opportunities in enablers like assurance tools, shared platforms, and talent programs.