America's AI Action Plan for Workers and State Laws
Its worker provisions focus on skills and labor research, while its state-law proposal concerns discretionary federal funding.
What the AI Action Plan Is and Is Not
America's AI Action Plan is a July 2025 White House policy roadmap with three pillars: innovation, infrastructure, and international diplomacy and security. It is not a statute or an executive order. Its recommendations address worker training and research, state AI rules, and other subjects, but the plan itself does not impose duties on companies, consumers or states.
The White House's America's AI Action Plan sets out recommended policy actions for agencies under those pillars. The plan's own text, retrieved November 17, 2026, is the source for every provision below. The plan was published in July 2025, so its recommendations should not be mistaken for a current list of completed agency actions as of November 17, 2026.
The plan calls its agency proposals "Recommended Policy Actions." A recommendation that an agency consider a state's rules does not, by itself, change a funding award. A recommendation to issue training guidance does not, by itself, make every training course eligible for tax-free assistance.
The plan's Introduction places American workers within its AI policy. Its worker proposals appear mainly under "Empower American Workers in the Age of AI," within the innovation pillar. The state-funding proposal appears under "Remove Red Tape and Onerous Regulation." One discusses preparation for labor-market change. The other discusses how federal agencies might treat state AI regulation.
Plan Provisions at a Glance
The White House plan assigns different recommended actions to different agencies. The table distinguishes each proposal from an action the plan itself has already taken.
| Provision | Lead agency | What the plan recommends | What it does not do |
|---|---|---|---|
| AI skills | Department of Labor, Department of Education, National Science Foundation and Department of Commerce | Prioritize AI skills in relevant education and workforce funding streams. | Give every worker a training place. |
| Training assistance | Department of the Treasury | Clarify when AI literacy and skills programs may qualify as educational assistance under Section 132 of the Internal Revenue Code. | Make every AI course tax-free. |
| Labor-market evidence | Bureau of Labor Statistics and Department of Commerce, through the Census Bureau and Bureau of Economic Analysis | Study AI adoption, jobs, displacement and wages using existing data. | Report a finding about how many jobs AI has displaced. |
| State regulatory climate | Office of Management and Budget, working with agencies | Have agencies consider state AI rules in AI-related discretionary funding decisions and limit funding in the circumstances described by the plan, consistent with applicable law. | Repeal a state law or automatically withhold an award. |
| State rules and communications law | Federal Communications Commission | Evaluate whether state AI regulations interfere with its Communications Act of 1934 responsibilities. | Declare every state AI rule invalid. |
| Deepfake evidence | National Institute of Standards and Technology and Department of Justice | Consider forensic guidance and a voluntary benchmark. Issue guidance to agencies exploring a deepfake evidence standard. | Create a new consumer prohibition. |
Each row summarizes the plan's "Recommended Policy Actions." None is a verified outcome. In particular, the state-funding proposal expressly includes "consistent with applicable law."
What It Recommends for Workers
The plan's "Empower American Workers in the Age of AI" section recommends skills funding, research and retraining. It does not set a limit on employers' use of AI. For the wider question of how AI may affect work, see AI and jobs.
The skills proposal asks the Department of Labor, Department of Education, National Science Foundation and Department of Commerce to make AI skill development a core objective of relevant funding streams. The plan names career and technical education, workforce training and apprenticeships. That is a direction for agencies administering those streams. The plan does not identify a course a worker can enroll in or guarantee an individual a place.
The Treasury proposal concerns an existing category of educational assistance. Under "Empower American Workers in the Age of AI," the plan asks the Department of the Treasury to clarify that many AI literacy and skills programs may qualify under Section 132 of the Internal Revenue Code. The plan says the guidance would enable employers to offer tax-free reimbursement for AI-related training in applicable situations.
The research proposals ask the Bureau of Labor Statistics and the Department of Commerce, through the Census Bureau and Bureau of Economic Analysis, to examine AI's effect on the labor market. The plan names AI adoption, job creation, displacement and wage effects as subjects for study. It points to existing data, including the Census Bureau's Business Trends and Outlook Survey. Those are questions for the proposed research. They are not findings the plan reports.
The same section proposes an AI Workforce Research Hub under the Department of Labor. Its job would be recurring analysis and scenario planning about AI's labor-market effects. The hub is proposed as a way to learn what is happening. The plan's separate retraining item concerns available discretionary funding.
For displaced workers, the plan recommends that the Department of Labor use available discretionary funding for rapid retraining. It also calls for guidance on identifying eligible dislocated workers and on how state Rapid Response funds can proactively upskill workers at risk of displacement. The plan proposes Department of Labor and Department of Commerce pilots carried out by states and workforce intermediaries. They would run under the Workforce Innovation and Opportunity Act and the Public Works and Economic Development Act. These are proposed routes for agency action. The roadmap creates no new benefit.
Worker training appears in the infrastructure pillar too. Under "Train a Skilled Workforce for AI Infrastructure," the plan proposes a Department of Labor and Department of Commerce initiative to identify high-priority occupations for AI infrastructure construction. It names electricians and advanced heating, ventilation and air conditioning technicians as examples. Voluntary skill frameworks, industry-driven training, early exposure and expanded Registered Apprenticeships are among the recommendations. This infrastructure proposal concerns jobs building AI infrastructure, a different task from studying displacement across the labor market.
What It Says About State AI Laws
The plan's "Remove Red Tape and Onerous Regulation" section asks federal agencies to weigh state AI regulatory climate in certain funding decisions, consistent with applicable law. The Office of Management and Budget is directed in the recommended action to work with agencies that administer AI-related discretionary funding. Those agencies would consider a state's regulatory climate and limit funding if its AI rules may hinder the effectiveness of the funding or award, "consistent with applicable law."
The same section says the federal government should not interfere with states' rights to pass prudent laws that are not unduly restrictive to innovation. The text expresses both a preference against rules it considers burdensome and a stated place for state lawmaking. It does not specify that every state AI law will receive the same treatment.
A separate recommended action asks the Federal Communications Commission to evaluate whether state AI regulations interfere with its obligations and authorities under the Communications Act of 1934. An instruction to evaluate a question is different from a finding that a named state law interferes. The plan names no state rule in that recommended action as already invalid.
The section also proposes a request for information on federal rules that hinder AI, work on federal rules the plan considers unnecessary, and a review of certain Federal Trade Commission investigations begun under the previous administration. These proposals should not be folded into a claim that the plan has already changed state law.
For the separate dispute over federal preemption, read state vs federal AI regulation. That page also covers a later executive order on state AI laws. The US AI regulation overview places this plan among broader policy developments.
Deepfake Evidence in the Legal System
The plan's "Combat Synthetic Media in the Legal System" section focuses on the handling of possible deepfake evidence. It notes that the TAKE IT DOWN Act had already been signed and says further work is needed because fake evidence could affect plaintiffs and defendants. The plan's proposed next steps concern forensic evaluation and evidence practice. The plan itself imposes no new rule on consumers.
One recommended action asks the National Institute of Standards and Technology, within the Department of Commerce, to consider developing its Guardians of Forensic Evidence evaluation program into a formal guideline. The proposal also calls for a companion voluntary forensic benchmark. The proposed benchmark is voluntary and is not described as a compulsory test for the public.
Another asks the Department of Justice to guide agencies that conduct adjudications as they explore a deepfake evidence standard. The plan points to a proposed Federal Rules of Evidence provision under consideration by the Advisory Committee on Evidence Rules. It also recommends that the Department of Justice's Office of Legal Policy comment on proposed deepfake-related additions to those rules. None of those recommendations says that the plan itself has adopted a new evidence rule. For existing legislation, see deepfake laws.
Where the Plan Stands Today
As of November 17, 2026, the plan's text shows what the White House recommended in July 2025. It does not show whether each agency later carried out those recommendations. The source for the provisions above is the plan itself. Its age is material when reading a proposed research hub, guidance or funding limit as though it were a current program.
The worker section proposes Treasury guidance. It does not supply that guidance. The state section proposes a funding consideration, subject to applicable law. It does not identify an award withheld under the proposal. The deepfake section proposes work on forensic and evidence standards. It does not enact those standards.
This also answers the executive-order question. The AI Action Plan is a roadmap, even though its Introduction refers to earlier executive orders. A later executive order on state AI laws is covered separately in the state and federal AI regulation explainer. Its existence does not turn the July 2025 plan into an executive order.
Our Takeaway
The White House plan states policy priorities, and its "Recommended Policy Actions" are proposals for agency work. For workers, the concrete proposals concern skills, research and retraining, with a separate training track for AI infrastructure jobs. For states, the proposal is consideration of AI rules in certain discretionary funding decisions, consistent with applicable law, plus a review by the Federal Communications Commission. The plan itself neither changes a state law nor creates a consumer obligation.
What Would Change Our Answer
Agency action or legislation could change the practical answer while leaving the plan's text unchanged. Guidance on training assistance would clarify the Treasury proposal. Agency decisions about AI-related discretionary funding would give the state-climate proposal a concrete application. Legislation on AI preemption or workforce funding could establish rules beyond the plan's recommendations. As of November 17, 2026, those later outcomes have not been established by the plan text.
Read the "Empower American Workers in the Age of AI" and "Remove Red Tape and Onerous Regulation" sections in the White House AI Action Plan to compare its worker and state-law recommendations directly.