US Federal AI Policy Shifts to Execution

Quick Take: The federal government is moving from AI governance frameworks to large‑scale deployment, earmarking billions in R&D and procurement through 2025 to accelerate adoption across agencies.

The United States has turned a page on AI policy: after years of drafting ethical guidelines, the current administration is funding and mandating concrete AI projects in defense, health, and civilian services. The shift is codified in the America’s AI Action Plan and reinforced by a series of executive orders signed in July 2025 that require agencies to submit AI use‑cases, adopt risk‑management frameworks, and report spend quarterly.

Funding Surge: Billions Flow Into Federal AI R&D

Federal AI and IT research spending jumped by $2.8 billion between FY 2021 and FY 2025, reaching a cumulative $31 billion in the five‑year window. The National Institute of Standards and Technology’s (NIST) AI Risk Management Framework, released in 2023, now underpins a substantial portion of the FY 2025 AI R&D request from NITRD agencies. This financing fuels both defense‑only and non‑defense programs, with the non‑defense share slated to double under the latest White House recommendation.

From Governance to Execution: New Mandates for Agencies

The Brookings analysis notes that the Trump administration (2025‑2026) has issued three executive orders that translate high‑level policy into actionable procurement rules, data‑governance standards, and workforce training requirements. Agencies must now:

1. Publish AI inventories and risk assessments per the OMB M‑24‑10 memorandum.
2. Adopt the NIST AI RMF for every new system, ensuring transparency, fairness, and accountability.
3. Report quarterly spend on AI projects to the Office of Management and Budget, creating a real‑time dashboard of federal AI investment.

These steps aim to close the “implementation gap” that earlier reports identified as a major barrier—poor data quality, siloed governance, and lack of skilled personnel.

Implementation Hurdles Remain

Despite the funding surge, challenges persist. A 2025 OECD study highlights that poor data governance still leads to sub‑optimal outcomes in several agencies, risking bias and inefficiency. Moreover, the state‑level legislative boom—all 50 states plus territories introduced AI bills in 2025—creates a patchwork of compliance requirements that federal agencies must navigate when sharing data with state partners.

State & Local Legislation: A Patchwork of Rules

The National Conference of State Legislatures reports that over 120 AI‑related bills were filed in 2025, ranging from biometric privacy to algorithmic transparency. While the federal government pushes for unified standards, the divergent state laws could complicate cross‑jurisdictional AI deployments, especially in health‑care and law‑enforcement domains.

What It Means for Israel

Israel’s AI ecosystem, backed by the Israel Innovation Authority, mirrors the U.S. trend of moving from policy to execution. Using the typical Israeli automation cost figures, a government department that automates a 10‑hour‑per‑week support task (≈ 1,560 hours / year) could save ≈ 936 hours annually (⁦60%⁩ automatable). At a loaded cost of ₪90 / hour, the annual savings equal ₪84,240. Building a medium‑complexity solution costs about ₪45,000 one‑time, delivering a payback in roughly 6.4 months—a compelling ROI that aligns with the U.S. emphasis on rapid, measurable outcomes.

For Israeli agencies, the U.S. model offers a blueprint: secure dedicated R&D budgets, adopt a risk‑management framework (e.g., NIST AI RMF), and institute transparent reporting. By doing so, they can tap into existing federal‑level expertise, accelerate AI pilots, and demonstrate fiscal responsibility to local stakeholders.

Forward Look: Toward a Coordinated AI Future

The next few years will test whether the U.S. can sustain its substantial investment and translate it into tangible public‑service improvements. Success will depend on continuous funding, robust data governance, and harmonized state‑federal regulations. Israel, watching closely, can leverage these lessons to shape its own AI rollout, ensuring that policy translates swiftly into execution and measurable public benefit.

Sources & further reading

• Original source: Google News — policy
• AMERICA'S AI ACTION PLAN | The White House
• July 2025 AI Developments Under the Trump Administration
• Artificial Intelligence (AI) - United States Department of State
• From governance to execution in federal AI policy | Brookings
• Summary of Artificial Intelligence 2025 Legislation