Skadden, Arps, Slate, Meagher & Flom LLP and Affiliates

White House Releases AI Action Plan: Key Legal and Strategic Takeaways for Industry

Skadden Publication / AI Insights

Stuart D. Levi David A. Simon Brian J. Egan Ken D. Kumayama Michael E. Leiter William E. Ridgway Katie Clarke Joshua Silverstein Tatiana O. Sullivan Melissa Muse

Executive Summary

  • What is new: The Trump administration’s AI Action Plan reflects a striking shift in approach, with the federal government driving development, expansion and regulation, focusing on deregulation, permitting, procurement and export control to ensure American domination of AI technologies.
  • Why it matters:The plan promises to provide a wide range of business opportunities for AI developers and supporting industries such as data centers, while also providing support for AI training across sectors.
  • What to do next: Businesses will want to study the proposals for opportunities as well as new requirements, and monitor the government’s planned RFIs, which will provide an opportunity to influence the details of the new policies.

__________

On July 23, 2025, the White House unveiled its long-awaited artificial intelligence (AI) policy blueprint, “Winning the AI Race: America’s AI Action Plan” (Plan), issued pursuant to Executive Order 14179, “Removing Barriers to American Leadership in Artificial Intelligence.” The Plan outlines over 90 near-term federal actions intended to reshape the U.S. AI landscape with strategic implications across the public and private sectors.

The Plan sets forth a sweeping vision for U.S. leadership in artificial intelligence, and is structured around three core pillars:

(I) Accelerating AI innovation.

(II) Building domestic AI infrastructure.

(III) Leading in global AI diplomacy and security.

The Plan is animated by several unifying principles that cut across each domain, including:

  • Placing American workers at the center of its AI strategy by advancing automation and infrastructure that will expand opportunities and raise living standards, not replace human labor.
  • Requiring that federally procured advanced “frontier” models reflect “objective truth” rather than what the Plan characterizes as “top-down ideological bias[es]” that fail to protect freedom of expression.
  • Implementing constant vigilance against foreign adversaries seeking to exploit U.S. AI technologies, by emphasizing export controls, intellectual property protection and national security screening as essential tools to secure the AI ecosystem.

Key Points

The AI Action Plan reflects a striking shift in approach to AI development, expansion and regulation at the federal level in the U.S. It stands in stark contrast to the more cautious approach advanced by the Biden administration, focusing instead on innovation and federal action across deregulation, permitting, procurement and export control to enable rapid adoption of U.S. AI technologies. Below, we highlight some of the most salient takeaways for businesses to better understand what the AI Action Plan means for them:

Expanded government procurement opportunities. The Plan calls for agencies to identify and eliminate regulatory and procurement barriers to AI innovation, particularly those that slow the government’s ability to adopt new tools. Public and private actors should prepare for Requests for Information (RFIs) on these topics and consider how their organizations might participate in shaping workforce upskilling initiatives that the administration is expected to fund.

Focus on physical infrastructure to support AI advancements. AI development is now explicitly tied to physical infrastructure, with a strong emphasis on accelerating development and construction of data centers, modernizing the energy grid and expanding high-performance computing capacity. Stakeholders in energy, utilities and real estate should evaluate how to position for expedited permitting processes — including potential categorical exclusions — and identify federal grants and loan programs that may support build-out.

New AI export opportunities. The Plan signals a new focus on exporting “full-stack AI technology packages,” integrating U.S.-developed software, hardware and technical support to compete globally. This initiative could bring significant commercial benefits to American companies. At the same time, export controls and outbound investment reviews continue to tighten for the technology to support high-end AI processing and computations (“compute”) and semiconductor technologies. Companies involved in cross-border transactions involving such advanced technologies should assess compliance strategies and consider engaging in shaping future trade-related AI policies.

AI model developers continue to face uncertainty. It is not yet clear how the administration will determine if a model is “objective and free from top-down ideological bias.” For now, entities that market their AI models to the government should consider how they might prove that their models adhere to this mandate. The Plan also signals a strong preference for “open-source and open-weight AI” as a means to foster innovation and create U.S. global standards. Model developers should consider how the administration’s position might impact their own “proprietary versus open” decisions regarding source code, weights, training data and other technical information.

Copyright and training data not discussed. The Plan does not address the contested issue of using copyrighted materials to train AI models, despite calls from some developers for the Plan to declare such training to be fair use in order to promote innovation. Instead, the Plan is silent on any copyright-related issues, indicating the administration views the issue as one for judicial interpretation. Nonetheless, President Trump made his personal sentiments on the issue clear when announcing the Plan, stating, “You can’t be expected to have a successful AI program when every single article, book or anything else that you’ve read or studied, you’re supposed to pay for. You just can’t do it, because it’s not doable.”

Support for widespread private sector AI integration. The Plan includes a number of proposals for fostering a “try-first” culture for AI across industries. These proposals include domain-specific efforts to accelerate adoption and regulatory “sandboxes.” Companies should monitor developments in their respective sectors to respond and be prepared to take advantage of specific opportunities to shape the ongoing AI implementation dialogue through Requests for Information and other initiatives developed out of the Plan.

Below, we describe in more detail the priorities and directives set forth in the Plan, as well as the anticipated implications for stakeholders across various sectors.

Pillar I: Accelerate AI Innovation

The first pillar prioritizes broad deregulatory efforts aimed at catalyzing private sector innovation while leveraging federal procurement and investment tools to steer AI development.

Deregulation and Open-Source Support

  • Federal funding for AI-related projects will be steered away from states with burdensome AI regulations, and the Federal Communications Commission will evaluate whether state-level AI rules interfere with federal authority. We note that this statement comes on the heels of Congress’ rejection of an amendment to the so-called “One Big Beautiful Bill Act” that would have imposed a ten-year moratorium on state AI laws. The Plan itself notes that it does not seek to hamper the right of states to enacted “prudent” AI laws that do not hamper innovation.
  • The Plan tasks the Federal Trade Commission with reviewing and terminating any existing investigations, final orders or consent decrees that are found to “advance theories of liability that unduly burden AI innovation.”
  • The Plan strongly encourages the development and adoption of open-source and open-weight AI models, recognizing their value for startups, academic research and government use, and their importance in driving U.S.-developed AI models as global standards. Recommended actions include improving the financial market for compute (e.g., spot and forward markets for AI compute resources), expanding access to private sector computing through the National AI Research Resource pilot under the auspices of the National Science Foundation (NSF), and convening stakeholders to drive open-source adoption among small and medium-sized businesses.
  • The Plan tasks the Department of Commerce (DOC) with facilitating public-private dialogue to identify supply chain issues that have hindered the use of AI in robotics and drone manufacturing, especially those with autonomous capabilities that can have defense applications.

Government and Military Use

  • The Plan calls for the comprehensive review and removal of federal procurement regulations that hinder AI innovation and adoption. The Office of Science and Technology Policy (OSTP) is tasked with launching an RFI from businesses and the general public to identify such barriers, while the Office of Management and Budget (OMB) will work with agencies to revise or repeal unnecessary rules.
  • The Plan centralizes all interagency collaboration on AI adoption efforts under the Chief Artificial Intelligence Officer Council, with OMB and the General Services Administration managing an AI procurement toolbox that will allow for uniform access to AI systems across the federal government.
  • OMB will lead a cohort of agencies with “High Impact Service Providers” tasked with increasing AI adoption to improve services provided to the public.
  • The Department of Defense (DoD) is tasked with identifying which workflows involved in major operational and enabling functions can be permanently transitioned to an automated AI process.
  • The Plan calls for DoD’s agreements with certain key private sector entities, such as cloud service providers and operators of computing infrastructure, to include terms providing for the government’s “priority access” to computing resources in the case of a national emergency or other significant conflict.

Workforce and Education

  • The Plan advances a “worker-first” agenda, with the Department of Labor (DOL), Department of Education (ED), NSF and DOC prioritizing AI skill development in education and workforce funding streams. This includes integrating AI into career and technical education, workforce training and apprenticeships.
  • The Department of the Treasury will issue guidance to clarify that AI literacy and skill development programs may qualify as eligible educational assistance, enabling employers to “offer tax-free reimbursement for AI-related training.”
  • The Bureau of Labor Statistics, Census Bureau and Bureau of Economic Analysis will “study AI’s impact on the labor market,” including impacts on job creation, worker displacement and wages.
  • A new AI Workforce Research Hub will be established to evaluate AI’s impact on the labor market and inform policy, while rapid retraining programs and pilot initiatives will be launched to help workers adapt to AI-driven changes.

Ideological Bias in AI Models

  • The Plan also advances the efforts of the recent White House Executive Order, “Preventing Woke AI in the Federal Government,” which highlighted the administration’s efforts to ensure ideological neutrality in federally-procured AI systems.
  • The Plan states that AI systems must be “free from ideological bias and be designed to pursue objective truth rather than social engineering agendas.” It recommends revising the National Institute of Standards and Technology (NIST) AI Risk Management Framework to “eliminate references to misinformation, Diversity, Equity, and Inclusion, and climate change,” and updating federal procurement guidelines to require objectivity in AI systems the government procures. The Plan does not define “objectivity” and developers with federal contracts or sales opportunities should closely track this space for further developments that clarify federal expectations. Given the complexities of defining, measuring and ensuring neutrality in AI systems, defining “objectivity” will likely generate substantial controversy.
  • In addition, the Plan calls on NIST and the Department of Justice to take further action, including proposed updates to the Federal of Rules of Evidence, to address non-consensual deepfakes on the internet.

Pillar II: Build American AI Infrastructure

The second pillar focuses on developing the physical foundations of the U.S. AI ecosystem, spanning energy, data and semiconductor infrastructure.

AI Infrastructure Buildout

  • The administration seeks to streamline permitting for data centers, semiconductor manufacturing and energy infrastructure by establishing new Categorical Exclusions under the National Environmental Policy Act (NEPA) and expanding the use of the FAST-41 process for eligible projects.
  • Federal lands will be made available for data center and power infrastructure construction and security guardrails will be implemented to ensure that AI infrastructure is free from adversarial technology.
  • The Plan also includes a national initiative to identify “high-priority occupations” that are essential to building AI infrastructure. Multiple agencies are also expected to collaborate on expanding early career exposure and registered apprenticeships in these fields.
  • The Plan prioritizes stabilizing and optimizing the U.S. electric grid to meet the demands of AI, including preventing premature decommissioning of power generation resources, upgrading transmission systems and embracing new energy sources such as nuclear and geothermal.
  • Revitalizing American semiconductor manufacturing is a central goal, with the DOC focusing on delivering a strong return on investment for CHIPS-funded projects and removing extraneous policy requirements. There is also an emphasis on integrating advanced AI tools into semiconductor manufacturing processes.
  • The Plan calls for building high-security data centers for military and intelligence use, advancing agency adoption of classified compute environments and creating technical standards for secure AI data centers.

Security of AI-Enabled Infrastructure

  • Critical infrastructure operators should prepare for new AI-specific cybersecurity expectations, including expanded use of model-specific threat simulations (“red-teaming”), and integration of AI into the Cybersecurity Performance Goals coordinated by the Department of Homeland Security and sector-specific agencies.
  • AI developers and deployers — particularly in regulated sectors — are encouraged to adopt “secure-by-design” principles, including pre-deployment safety testing, access controls, traceability mechanisms and fail-safe features to mitigate model misuse, extraction or manipulation.
  • Federal agencies will coordinate the development of AI-tailored incident response protocols, with playbooks addressing detection, reporting, escalation and containment of AI system failures—including hallucinations, adversarial prompts and poisoning attacks.
  • NIST will lead the creation of voluntary cybersecurity and safety benchmarks for AI systems, which are expected to influence future agency procurement standards and could become de facto expectations in grant and regulatory contexts.
  • Companies operating AI in critical infrastructure environments should assess whether current enterprise risk and incident response plans account for AI-specific failure modes, including model behavior drift, data integrity attacks and unsafe autonomous system actions

Pillar III: Lead in International AI Diplomacy and Security

The third pillar targets international AI competition — particularly with China — and emphasizes the strategic export of U.S.-developed AI systems and standards to allied and partner nations.

International Leadership and Security

  • The Plan calls on the U.S. to export its full AI technology stack — hardware, models, software, applications and standards — to “America’s AI Alliance,” with the DOC coordinating industry consortia and facilitating deals that meet U.S. security requirements.
  • The Plan directs the U.S. to counter Chinese influence in international AI governance bodies and advocate for innovation-friendly, American values-based standards.
  • Strengthening export controls on advanced AI compute and semiconductor manufacturing subsystems is a priority, including leveraging location verification features and expanding end-use monitoring to prevent diversion to adversaries.
  • The administration will work to align export controls globally, using diplomatic tools and secondary tariffs if necessary, and develop a technology diplomacy strategic plan to ensure allies adopt complementary AI protection systems.
  • The Plan also calls for evaluating the most advanced “frontier” AI systems for national security risks, recruiting leading AI researchers to federal agencies, and building national security-related AI evaluation capabilities.

Looking Ahead

America’s AI Action Plan sets an ambitious and detailed course for U.S. AI policy, emphasizing deregulation, infrastructure expansion, workforce development and international leadership as core strategic priorities. While many of the proposed initiatives will require agency-level implementation, intergovernmental coordination and possible legislative engagement, the Plan signals a clear federal posture: to accelerate AI advancement by minimizing regulatory friction and asserting U.S. dominance across the AI value chain.

Timeline: RFIs and Implementation

In addition to key actions outlined in the AI Action Plan itself, including forthcoming RFIs, several executive orders have been issued to implement the Plan’s broader recommendations and directives.

  • RFIs on regulatory barriers. OSTP is expected to issue an RFI seeking input on federal regulations hindering AI innovation.
  • Procurement guidance. OMB is directed to issue procurement implementation guidance for AI models within 120 days. (See “Preventing Woke AI in the Federal Government.”)
  • Infrastructure permits. An executive order mandates expedited permitting for qualifying data center and energy infrastructure projects within 180 days. (See “Accelerating Federal Permitting of Data Center Infrastructure.”)
  • American AI exports program. DOC must establish the American AI Exports Program, solicit industry consortia proposals for “full-stack” AI technology export packages, and evaluate submissions on a rolling basis for designation as priority AI export packages eligible for federal support and incentives. (See “Promoting the Export of the American AI Technology Stack.”)
  • We will continue to monitor regulatory developments, agency actions and related legal frameworks as further guidance emerges.

Associates Mana Ghaemmaghami, Abigail Hall, Lisa Rechden and Lisa Zivkovic contributed to this article.

See the Executive Briefing publication

This memorandum is provided by Skadden, Arps, Slate, Meagher & Flom LLP and its affiliates for educational and informational purposes only and is not intended and should not be construed as legal advice. This memorandum is considered advertising under applicable state laws.

BACK TO TOP