America's AI Action Plan
United States · Strategy · The White House
High impact. A reviewer confirms each change before it is listed here.
Our summary
A comprehensive national strategy outlining the U.S. government's approach to achieving global dominance in artificial intelligence development and deployment. The plan focuses on three main areas: accelerating innovation by removing regulatory barriers and promoting AI research, building critical infrastructure including semiconductors and data centers, and leading international diplomacy to ensure American AI standards worldwide. It emphasizes protecting American workers, ensuring AI systems reflect free speech values, and maintaining national security.
Written by the Index in our own words. It is not legal advice; the official text is the authority.
Inside this policy
Every section of this policy in order. Our notes summarise each one in plain language and do not quote the official text, which is the authority.
Drafted by the Index An automated summary of the official text, not yet reviewed by a person.Reviewed Checked by a person against the official text.
34 sections
Jump to a chapter
Introduction
America's AI Action Plan aims to advance U.S. AI leadership through three pillars: innovation with reduced regulatory barriers, vast AI infrastructure powered by energy resources, and establishing American AI as the global standard with allied cooperation. The Plan centers AI workers and beneficiaries in policy, requires systems free from ideological bias that pursue objective truth and provide trustworthy factual information, emphasizes AI should complement rather than replace American workers, and calls for protecting advanced technologies from misuse and theft while monitoring emerging or unforeseen risks.
Pillar I Accelerate AI Innovation
15 sections
Chapter covering the first pillar: supporting private-sector innovation to develop the world's most powerful AI systems and their transformative applications. Topics include removing regulatory barriers, protecting free speech and American values in AI systems, supporting open-source models, adoption incentives, worker support, manufacturing, science applications, datasets, interpretability research, evaluations, government adoption, defense applications, and protection against synthetic media misuse.
Remove Red Tape and Onerous Regulation
Federal agencies must solicit business and public input on regulations hindering AI innovation and take action; identify and revise regulations through OMB working with all agencies; consider state AI regulatory climate in discretionary AI funding decisions; have FCC evaluate whether state AI rules interfere with Communications Act obligations; and review FTC investigations and orders to ensure they do not unduly burden innovation.
Ensure that Frontier AI Protects Free Speech and American Values
Requires that advanced AI systems prioritize freedom of expression and reflect factual accuracy rather than ideology, with the government contracting only with developers of large language models that meet these standards. Actions include revising NIST's AI framework to remove specified topics, updating procurement standards, and evaluating Chinese frontier models for political alignment.
Encourage Open-Source and Open-Weight AI
Government should support open-source and open-weight models as beneficial for innovation, business, government adoption of AI, and academic research. Federal actions include improving compute financial markets for startups and academics, partnering with technology companies to increase research access to resources through NAIRR pilot, building sustainable NAIRR operations, publishing new AI R&D Strategic Plan, and convening stakeholders to drive small and medium business adoption.
Enable AI Adoption
Addresses barriers to AI adoption in established industries and proposes federal coordination to create a testing culture. Calls for regulatory sandboxes and centers of excellence where researchers and enterprises can deploy AI; domain-specific efforts to develop AI standards across healthcare, energy, and agriculture; ongoing assessments of US versus competitor AI adoption in national security; and collection of intelligence on foreign frontier AI projects with security implications.
Empower American Workers in the Age of AI
Supports worker-focused AI policies by directing agencies to prioritize AI skill development in education and workforce programs, with Treasury clarifying tax treatment of employer-funded AI literacy training, labor statistics agencies studying AI's impact on jobs and wages, and establishment of an AI Workforce Research Hub. Authorizes DOL to fund rapid retraining for workers displaced by AI and directs DOL and DOC to pilot new workforce approaches through states and intermediaries to identify scalable strategies for adapting to AI-driven labor market change.
Support Next-Generation Manufacturing
Prioritizes federal investment in manufacturing technologies for AI-enabled innovations including autonomous drones, robotics, and self-driving vehicles. Directs multiple agencies to invest in foundational and translational manufacturing through various programs and authorities; convenes industry and government stakeholders via NTIA to address robotics and drone manufacturing supply chain challenges.
Invest in AI-Enabled Science
Commits Federal agencies to fund automated laboratory infrastructure and research organizations that use AI to accelerate scientific discovery, incentivizes public data sharing by researchers, and requires federally funded researchers to disclose datasets used in AI model training.
Build World-Class Scientific Datasets
Establishes data as a strategic asset for AI innovation while protecting privacy and civil liberties. Directs NSTC to recommend minimum data quality standards for scientific datasets; promulgates OMB regulations on federal data accessibility; creates secure compute environments in NSF and DOE for controlled access to restricted data; establishes an online portal for NSF's data service; and explores a whole-genome sequencing program on federal lands to support biological AI models.
Advance the Science of AI
Prioritizes investment by the United States in foundational research across theory, computation, and experimentation to discover novel AI paradigms that enhance capabilities, reflecting this direction in the forthcoming National AI Research and Development Strategic Plan.
Invest in AI Interpretability, Control, and Robustness Breakthroughs
The United States will fund research programs to understand how frontier AI systems work, predict their behavior, and make them more secure and reliable for use in high-stakes national security applications. Key actions include launching a technology development program through DARPA in collaboration with CAISI and NSF, prioritizing fundamental advances in AI interpretability and control in the National AI R&D Strategic Plan, and coordinating an AI hackathon led by DOD, DOE, CAISI, DHS, and NSF with academic partners to test AI systems for transparency, effectiveness, control, and security vulnerabilities.
Build an AI Evaluations Ecosystem
Directs NIST at DOC to publish guidelines for Federal agencies to conduct AI evaluations for their missions and legal compliance. Directs NIST, DOE, NSF, and other Federal agencies to support developing science of measuring and evaluating AI models. Requires CAISI-hosted meetings at least twice yearly for sharing evaluation learnings.
Accelerate AI Adoption in Government
Formalize the Chief Artificial Intelligence Officer Council as primary venue for interagency AI coordination. Create a talent-exchange program for rapid details of Federal staff with specialized AI skills. Establish an AI procurement toolbox managed by GSA with OMB coordination, allowing agencies to choose among multiple models compliant with privacy, data governance, and transparency laws.
Drive Adoption of AI within the Department of Defense
Directs DOD to identify required AI talent and workforce skills, then implement development programs. Directs DOD to establish an AI and Autonomous Systems Virtual Proving Ground, beginning with scoping technical and security requirements. Directs DOD to develop streamlined processes for classifying and optimizing workflows for AI automation.
Protect Commercial and Government AI Innovations
The United States government will work with American AI developers to protect their innovations and systems from security threats, including cyberattacks, insider threats, and other risks, while balancing the benefits of deploying cutting-edge AI technologies with national security requirements. This effort is led by DOD, DHS, CAISI, and members of the Intelligence Community.
Combat Synthetic Media in the Legal System
Federal agencies will develop tools and guidance to address the risks that AI-generated deepfakes pose to the courts and law enforcement, particularly the risk of fake evidence denying justice to plaintiffs or defendants. Actions include NIST developing a formal deepfake evaluation guideline and voluntary forensic benchmark, DOJ issuing guidance to agencies conducting adjudications on adopting a deepfake evidence standard, and DOJ's Office of Legal Policy submitting formal comments on any deepfake-related changes to the Federal Rules of Evidence.
Pillar II Build American AI Infrastructure
8 sections
This pillar addresses AI infrastructure development given stagnant American energy capacity since the 1970s compared to China's grid growth. It covers streamlined permitting for data centers, semiconductors, and energy infrastructure with security; grid development matching AI innovation pace; domestic semiconductor manufacturing; secure data centers for military and intelligence; skilled workforce training; critical infrastructure cybersecurity; secure-by-design technologies; and federal AI incident response capacity.
Create Streamlined Permitting for Data Centers, Semiconductor Manufacturing Facilities, and Energy Infrastructure while Guaranteeing Security
The strategy calls for expediting environmental permitting for data center and semiconductor manufacturing facilities through establishing categorical exclusions under NEPA, expanding the FAST-41 process, exploring a nationwide Clean Water Act permit for data centers, and streamlining regulations under the Clean Air Act and Clean Water Act. It also directs federal agencies to make land available for data center and power generation construction.
Develop a Grid to Match the Pace of AI Innovation
Requires developing a comprehensive grid strategy across four elements: stabilizing existing assets and preventing premature decommissioning of critical power generation; optimizing transmission efficiency and demand management; prioritizing interconnection of reliable power sources and embracing new energy generation technologies; and creating a long-term strategic blueprint for grid expansion aligned with AI infrastructure needs.
Restore American Semiconductor Manufacturing
Directs the Department of Commerce's CHIPS Program Office to focus on return on investment for taxpayers and remove extraneous policy requirements for CHIPS-funded semiconductor projects. Also requires reviewing semiconductor programs to ensure they accelerate integration of advanced AI tools into manufacturing.
Build High-Security Data Centers for Military and Intelligence Community Usage
AI systems will process sensitive government intelligence data, requiring data centers protected against advanced nation-state attacks. Development of new technical standards for secure AI data centers will be led by DOD, the Intelligence Community, NSC, and NIST, collaborating with industry and relevant research centers. Federal agencies should adopt classified computing environments for secure and scalable AI operations.
Train a Skilled Workforce for AI Infrastructure
Multiple agencies must collaborate to develop priority occupations in AI infrastructure, create skill frameworks, support industry-driven training programs co-developed with employers, expand early career exposure for middle and high school students, update career and technical education curricula, and streamline Registered Apprenticeship programs in critical infrastructure occupations.
Bolster Critical Infrastructure Cybersecurity
Directs Homeland Security to lead an information-sharing center for AI security threats across critical infrastructure, issue guidance on remedying AI vulnerabilities, and coordinate disclosure of known AI vulnerabilities from federal agencies to the private sector using existing mechanisms. Requires AI systems in safety-critical or security applications to be built with defensive capabilities, robustness, and instrumentation to detect performance degradation and malicious activities.
Promote Secure-By-Design AI Technologies and Applications
The government must ensure that AI systems used in national security applications are protected against malicious or spurious inputs such as data poisoning and privacy attacks. DOD will refine its Responsible AI and Generative AI Frameworks in collaboration with NIST and ODNI. ODNI will publish an Intelligence Community Standard on AI Assurance under Intelligence Community Directive 505.
Promote Mature Federal Capacity for AI Incident Response
Requires integrating AI considerations into incident response frameworks across public and private sectors. NIST will partner with industry to establish standards and response frameworks. CISA must modify cybersecurity playbooks to address AI systems and require Chief Information Security Officers to consult Chief AI Officers and other officials during incidents.
Pillar III Lead in International AI Diplomacy and Security
7 sections
Establishes a pillar focused on global AI competition, aiming to promote American AI systems, computing hardware, and standards worldwide while preventing adversaries from accessing U.S. innovation. Contains subsections on exporting AI, countering foreign influence in governance bodies, strengthening export controls, and biosecurity investments.
Export American AI to Allies and Partners
Directs DOC to establish a program gathering industry proposals for full-stack AI exports to aligned nations. DOC will coordinate with the Economic Diplomacy Action Group, Trade and Development Agency, Export-Import Bank, International Development Finance Corporation, and State Department to facilitate deals meeting U.S. security requirements and standards.
Counter Chinese Influence in International Governance Bodies
Directs State and Commerce Departments to use American influence in international forums to promote AI governance frameworks prioritizing innovation and American values while opposing authoritarian influence, particularly efforts by Chinese entities to shape surveillance and facial recognition standards.
Strengthen AI Compute Export Control Enforcement
DOC, OSTP, and NSC will explore using location verification features on advanced AI compute chips to prevent their use in countries of concern. DOC will lead enhanced monitoring of global chip exports and emerging AI compute developments to detect diversions, with expanded end-use monitoring in high-risk countries lacking in-country Export Control Officers.
Plug Loopholes in Existing Semiconductor Manufacturing Export Controls
DOC will develop new export controls on semiconductor manufacturing component subsystems. Currently, the U.S. and allies control major semiconductor manufacturing systems but not many component sub-systems that require regulation to prevent adversary access.
Align Protection Measures Globally
Directs multiple agencies to develop complementary technology protection measures, create a technology diplomacy plan for an AI alliance ensuring allies adopt aligned controls, expand plurilateral controls for the AI technology stack, and coordinate with allies on export controls, preventing adversaries from supplying their defense industrial bases.
Ensure that the U.S. Government is at the Forefront of Evaluating National Security Risks in Frontier Models
CAISI at DOC will evaluate advanced AI systems for national security risks including cyberattacks, CBRNE weapons development, and vulnerabilities, in partnership with frontier AI developers. CAISI will assess potential security vulnerabilities and foreign influence in adversary AI systems used in critical infrastructure. The Federal government will recruit leading AI researchers and maintain national security-related AI evaluations through interagency collaboration with research institutions.
Invest in Biosecurity
Requires all institutions receiving federal research funding to use nucleic acid synthesis tools with robust sequence screening and customer verification procedures, with enforcement mechanisms rather than voluntary compliance. OSTP will convene actors to develop data sharing mechanisms for screening fraudulent customers. Directs building national security AI evaluations through interagency collaboration.
Controls
What this policy requires, forbids, grants or penalises, one control at a time and in our own words. Each control links to the official text, which is the authority. This is not legal advice.
Drafted by the Index Drafted automatically from the official text, not yet reviewed by a person.Reviewed Checked by a person against the official text.
Themes covered 19 of 35
Choose a theme to see only its controls. About the themes
Risk and scope 1 of 4
- Risk classification, not covered
- Prohibited practices, not covered
- Impact assessment, not covered
Data and privacy 3 of 4
- Data subject rights and consent, not covered
Transparency and records 4 of 4
Technical safeguards 2 of 4
- Human oversight, not covered
- Bias and non-discrimination, not covered
Assurance and lifecycle 2 of 4
- Conformity assessment and certification, not covered
- Registration and notification, not covered
Rights and protected groups 0 of 4
- Automated decision-making rights, not covered
- Children and vulnerable groups, not covered
- Biometric identification and surveillance, not covered
- Whistleblowing, complaints and redress, not covered
Roles and organisation 6 of 6
Enforcement and sectors 1 of 5
- Sector rules: health, not covered
- Sector rules: finance and insurance, not covered
- Sector rules: employment, not covered
- Sector rules: education, not covered
The full list of controls could not be loaded, so only the first 40 are shown.
Remove Red Tape and Onerous Regulation 1 control
- ObligationApplies to: Public authorityDrafted by the Index
The Federal Trade Commission should review all prior investigations to ensure they do not advance liability theories that unduly burden AI innovation, and review all final orders and consent decrees to identify any that may hinder AI development.
Encourage Open-Source and Open-Weight AI 1 control
- ObligationApplies to: Public authorityDrafted by the Index
The federal government, through collaboration with industry and the National Institute of Standards and Technology, the Office of Science and Technology Policy, and the National Science Foundation's pilot program, should accelerate the development of a financial market for computing access to benefit startups and academics.
Enable AI Adoption 1 control
- ObligationApplies to: Public authorityDrafted by the Index
The National Institute of Standards and Technology should lead domain-specific efforts in sectors such as healthcare, energy, and agriculture to convene stakeholders and accelerate development of national standards for AI systems and measure productivity gains.
Empower American Workers in the Age of AI 4 controls
- ObligationApplies to: Public authorityDrafted by the Index
Multiple federal agencies should emphasize teaching and building AI skills as a central goal within relevant education and workforce development funding initiatives.
- Applies to: ProviderDrafted by the Index
Employers should be enabled to offer workers tax-free reimbursement for AI-related training to increase private investment in AI skills development.
- ObligationApplies to: Public authorityDrafted by the Index
The Bureau of Labor Statistics and the Commerce Department should analyze AI's effects on the labor market using data collection mechanisms already in place.
- ObligationApplies to: Public authorityDrafted by the Index
The Department of Labor should establish an AI Workforce Research Hub to conduct sustained evaluation of how AI affects employment and workers.
Support Next-Generation Manufacturing 1 control
- ObligationApplies to: Public authorityDrafted by the Index
The Commerce Department should bring together industry and government actors to identify obstacles to American robotics and drone manufacturing in the supply chain.
Invest in AI-Enabled Science 4 controls
- ObligationApplies to: Public authorityDrafted by the Index
Multiple federal agencies should fund automated laboratory infrastructure across scientific fields to enable AI-driven scientific discovery.
- ObligationApplies to: Public authorityDrafted by the Index
Federal agencies should use long-term agreements to support research organizations that use AI and emerging technologies to advance fundamental scientific discovery.
- ObligationApplies to: Public authorityDrafted by the Index
Federal agencies should encourage researchers to make high-quality datasets publicly available by considering the scientific datasets from prior work when reviewing new project proposals.
- ObligationApplies to: AnyoneDrafted by the Index
Researchers receiving federal funding should disclose non-proprietary and non-sensitive datasets used by AI models during research and experimentation.
Build World-Class Scientific Datasets 5 controls
- ObligationApplies to: Public authorityDrafted by the Index
The National Science and Technology Council should issue recommendations on the minimum quality standards that biological, materials, chemical, physical and other scientific data must meet before use in training AI models.
- ObligationApplies to: Public authorityDrafted by the Index
The Office of Management and Budget should adopt regulations to increase the presumption that federal data can be accessed and expand secure access methods to lower barriers to federal data use by statistical agencies while protecting confidential information.
- ObligationApplies to: Public authorityDrafted by the Index
The National Science Foundation and the Department of Energy should create secure computing environments that allow controlled access to restricted federal data for AI purposes.
- ObligationApplies to: Public authorityDrafted by the Index
The National Science Foundation should launch an online portal for its National Secure Data Service demonstration project to give the public and federal agencies access to AI use cases involving restricted federal data under controlled conditions.
- ObligationApplies to: Public authorityDrafted by the Index
The National Science and Technology Council and partner federal agencies should consider developing a whole-genome sequencing program for organisms on federal lands to generate training data for biological foundation models.
Advance the Science of AI 1 control
- ObligationApplies to: Public authorityDrafted by the Index
The federal government should prioritize funding for groundbreaking research in theoretical and computational science to help the United States remain the world's leader in discovering new AI capabilities.
Invest in AI Interpretability, Control, and Robustness Breakthroughs 3 controls
- ObligationApplies to: Public authorityDrafted by the Index
The Department of Defense, in partnership with the Commerce Department and National Science Foundation, should launch a research program to advance understanding of how AI systems work, how to direct their behaviour, and how to defend them against adversarial attacks.
- ObligationApplies to: Public authorityDrafted by the Index
The federal government should recognize advances in AI interpretability, control and resilience as priorities in its forthcoming National AI research and development strategy.
- ObligationApplies to: Public authorityDrafted by the Index
Federal science agencies, with academic partners, should coordinate a hackathon to have researchers test AI systems for how well they explain their reasoning, their practical utility, control mechanisms, and security risks.
Build an AI Evaluations Ecosystem 5 controls
- ObligationApplies to: Public authorityDrafted by the Index
The National Institute of Standards and Technology should produce guidance and tools to help federal agencies assess how well their AI systems work and whether they comply with existing laws.
- ObligationApplies to: Public authorityDrafted by the Index
Federal science agencies, led by the National Institute of Standards and Technology, should support research that establishes rigorous methods for measuring and assessing AI systems.
- ObligationApplies to: Public authorityDrafted by the Index
The Commerce Department should hold at least two meetings annually for government and research institutions to exchange insights and proven methods for evaluating AI systems.
- ObligationApplies to: Public authorityDrafted by the Index
The Department of Energy and National Science Foundation should fund development of testbeds where researchers can test new AI systems in controlled, realistic settings and bring them to market, encouraging diverse teams and covering agriculture, transportation and healthcare.
- ObligationApplies to: Public authorityDrafted by the Index
The Commerce Department should use the National Institute of Standards and Technology to convene the AI Consortium to define shared metrics and technical standards that will support the growth of reliable, interoperable AI.
Accelerate AI Adoption in Government 6 controls
- ObligationApplies to: Public authorityDrafted by the Index
The federal government should establish the Chief Artificial Intelligence Officer Council as the main forum where agencies coordinate and collaborate on adopting AI tools, working alongside other councils on management, data, technology, statistics and privacy.
- ObligationApplies to: Public authorityDrafted by the Index
The federal government should set up a staff mobility program enabling agencies to borrow AI specialists, such as data scientists and engineers, from other agencies.
- ObligationApplies to: Public authorityDrafted by the Index
The General Services Administration, in coordination with the Office of Management and Budget, should establish a shared system that allows federal agencies to select from multiple AI models while meeting applicable rules on confidentiality, data stewardship and openness, and to document other agencies' uses.
- ObligationApplies to: Public authorityDrafted by the Index
The federal government should create a program through the General Services Administration to quickly move advanced AI tools and lessons learned between agencies.
- ObligationApplies to: Public authorityDrafted by the Index
Federal agencies must ensure, to the greatest extent feasible, that staff whose duties would benefit from advanced language models have access to them and receive necessary training.
- ObligationApplies to: Public authorityDrafted by the Index
The Office of Management and Budget should convene a group of agencies that deliver high-value public services to test and expand the use of AI for delivering better services to citizens.
Drive Adoption of AI within the Department of Defense 8 controls
- ObligationApplies to: Public authorityDrafted by the Index
The Department of Defense should identify what skills and expertise its personnel need to deploy AI at scale.
- ObligationApplies to: Public authorityDrafted by the Index
The Department of Defense should create and run talent development initiatives to build its workforce capacity in AI.
- ObligationApplies to: Public authorityDrafted by the Index
The Department of Defense should establish a virtual testing ground for AI and autonomous systems, starting with assessment of technical, geographic, security and resource needs.
- ObligationApplies to: Public authorityDrafted by the Index
The Department of Defense should create a rapid process to identify, assess and enhance workflows in its operations and support functions in order to find priority candidates for AI-driven automation.
- ObligationApplies to: Public authorityDrafted by the Index
When the Department of Defense automates a workflow with AI, it should convert that workflow to the AI-based method as soon as feasible.
- ObligationApplies to: Public authorityDrafted by the Index
The Department of Defense should reach agreements with cloud and computing providers to guarantee it preferential access to computing capacity during a national crisis.
- ObligationApplies to: Public authorityDrafted by the Index
The Department of Defense should develop AI research, education, and workforce hubs in its senior military educational institutions, teaching AI competency to military personnel.
- ObligationApplies to: Public authorityDrafted by the Index
Military colleges should teach curricula in AI use, development and infrastructure administration to prepare military officers.
Change history
No changes recorded since we began monitoring this policy.