⚡ Quick Summary
Published by the Bipartisan House Task Force on Artificial Intelligence, this report presents the task force's key findings and recommendations on United States AI policy. Created by Speaker Johnson and Democratic Leader Jeffries on February 20, 2024, the task force comprises twenty-four members drawn from twenty committees and engaged with over one hundred experts through hearings and roundtables during 2024.
Seven principles frame the analysis: identify AI issue novelty, promote AI innovation, protect against AI risks and harms, empower government with AI, affirm the use of a sectoral regulatory structure, take an incremental approach, and keep humans at the centre of AI policy. Fifteen chapters run from Government Use, Federal Preemption of State Law, Data Privacy, National Security and Research, Development, and Standards through Civil Rights and Civil Liberties, Education and Workforce, Intellectual Property, Content Authenticity, Open and Closed Systems, Energy Usage and Data Centers, Small Business, Agriculture, Healthcare and Financial Services.
The deliverable is 66 key findings and 89 recommendations, offered as "a blueprint for future actions that Congress can take"; the transmittal letter gives the recommendation count as 85.
🧩 What’s Covered
The extracted pages include the front matter and the first five substantive chapters; the later chapters appear here only as table-of-contents entries and overview summaries.
- Front matter and framing: a table of contents, a transmittal letter to Speaker Johnson and Democratic Leader Jeffries, the task force's creation, membership and engagement with experts, seven framing principles, and an overview of findings and recommendations for all 15 chapters.
- Government Use: principles for responsible government AI use, OMB memorandum M-24-10, documentation categories proposed for transparency (data and metadata, software, model development, deployment and use), the NTTAA, NIST's FIPS and AI Risk Management Framework, legacy IT modernisation, federal cybersecurity, and OPM's AI workforce guidance; the overview lists ten recommendations.
- Federal Preemption of State Law: the Supremacy Clause, express and implied preemption, floors and ceilings illustrated by HIPAA, the E-SIGN Act and the GDPR, and FCC broadband and 5G litigation as an analog; one recommendation to commission a study of federal and state AI regulation.
- Data Privacy: training-data sourcing and web scraping, synthetic data and model collapse, six categories of privacy harm (physical, economic, emotional, reputational, discrimination, autonomy), the absence of a comprehensive federal privacy law alongside nineteen state laws; two recommendations on privacy-enhanced data access and technology-neutral privacy law.
- National Security: DOD programmes (DARPA's AI Next, Project Maven, the Chief Digital and AI Officer, ADVANA, Tradewinds, Open DAGIR), the NSCAI's 543 recommendations, the October 2024 national security memorandum, technical obstacles in data, compute, model protection and talent, and China's AI ambitions; four recommendations.
- Research, Development, and Standards: federal AI R&D spending of $2.9 billion in 2023 and an agency-by-agency budget table for FY2022–FY2024, the National AI Initiative Act and Initiative Office, the National AI R&D Strategic Plan, the National AI Research Resource, 27 AI Research Institutes with $500 million in investments, NSF's TIP Directorate and DOE's FASST initiative; twelve recommendations.
- Later chapters: the overview records findings and recommendations on civil rights and civil liberties, education and workforce, intellectual property, content authenticity, open and closed systems, energy usage and data centers, small business, agriculture, healthcare and financial services, including human involvement in consequential decisions and a risk-based approach to synthetic content.
- Appendices: task force members, task force events, key government policies, areas for future exploration, an overview of AI technology and definitional challenges of AI.
💡 Why it matters?
The report is the House task force's consolidated statement of United States federal AI policy direction, so it shows which problems Congress was advised to solve and with which instruments. It maps existing levers — OMB memorandum M-24-10, NIST's Federal Information Processing Standards and AI Risk Management Framework, GAO's accountability framework — and notes that the first iteration of the risk management framework "only sets the theoretical baseline" and is not a standards document. For anyone assessing U.S. AI obligations or selling AI into government, it sets out the sectoral regulatory structure the task force endorses and the areas left to study, including preemption and data privacy.
❓ What’s Missing
Only a partial text extraction was available, covering 80 of 273 pages, so the detailed chapters from Civil Rights and Civil Liberties onward and the appendices were seen only through the overview. Within what is present, the report states that it "is certainly not the final word on AI issues for Congress", leaves federal preemption to a proposed study, and says that further exploration of data privacy is warranted. Its recommendations are directional rather than drafted legislative text, and the transmittal letter's count of 85 recommendations conflicts with the 89 given in the overview.
👥 Best For
Congressional staff and policy analysts tracking U.S. federal AI priorities; agency AI programme and governance leads aligning internal work with OMB, NIST and GAO expectations; compliance and public-policy teams at vendors selling AI into government; and sector regulators or researchers mapping how one general-purpose technology is treated across domains.
📄 Source Details
Leading AI Progress: Policy Insights and a U.S. Vision for AI Adoption, Responsible Innovation, and Governance, published by the Bipartisan House Task Force on Artificial Intelligence (118th Congress); no author byline is printed. The document is in English and runs to 273 PDF pages, of which pages 1–80 were available for this review as a partial text extraction; the cover page is a placeholder template and no publication date, version number or self-referencing URL is printed. Page references use the extracted PDF pages.
