Trump Administration Unveils AI Action Plan with Implications for Innovation, Infrastructure, and Global Tech Competition

On July 23, 2025, the Trump Administration unveiled a comprehensive national action plan for artificial intelligence (“Action Plan”), accompanied by three Executive Orders that begin implementing key elements of the plan: Promoting the Export of American AI Technology Stack, Accelerating Federal Permitting of Data Infrastructure, and Preventing Woke AI in the Federal Government.

The Action Plan is divided into three pillars focused on accelerating artificial intelligence (“AI”) innovation, building AI infrastructure, and leading in AI diplomacy and security. Key elements of these pillars and the related Executive Orders that could have important implications for businesses include:

• Identifying and eliminating regulations that burden AI innovation;
• Streamlining and incentivizing infrastructure that supports AI, including data centers, semiconductor manufacturing, and energy infrastructure;
• Driving adoption of AI by the federal government as well as various industries (such as healthcare and energy);
• Promoting American AI technology abroad, including through a new export program, while continuing to prevent foreign adversaries from accessing advanced compute; and
• Emphasizing ideological neutrality and accuracy in large language model (“LLM”) responses.

Below, we summarize key provisions of the AI Action Plan and the accompanying Executive Orders.

AI Action Plan

Pillar 1: Accelerate AI Innovation

The first pillar of the Action Plan focuses on steps intended “to create conditions where private-sector-led innovation can flourish” to ensure that the United States will “lead the world in creative and transformative application of [AI] systems.”

Remove Red Tape and Onerous Regulation: The Action Plan emphasizes that cumbersome federal or state AI regulations should not impede the private sector. Accordingly, the Action Plan calls for the initiation of a request for information process for the public to identify regulations burdening AI innovation and that federal agencies review or repeal existing regulations and policies that “unnecessarily hinder AI development or deployment.” Further, the Action Plan provides for a review of Federal Trade Commission (“FTC”) investigations to ensure that its enforcement activities do not “advance theories of liability that unduly burden AI innovation,” and could lead the Administration to modify or set aside prior FTC actions. In addition to federal regulatory action, the Action Plan targets state actions, instructing federal agencies with AI-related discretionary funding programs to “consider a state’s AI regulatory climate when making funding decisions and limit funding if the state’s AI regulatory regimes may hinder the effectiveness of that funding or award.” The Federal Communications Commission (“FCC”) also is recommended to evaluate whether state AI rules interfere with the FCC’s ability to carry out its obligations and authorities.

Ensure that Frontier AI Protects Free Speech and American Values: The Action Plan, and related Executive Order on “Preventing Woke AI,” emphasize that “free speech flourishes in the era of AI and that AI procured by the Federal government objectively reflects truth rather than social engineering agendas.” To support this stated goal, the Action Plan recommends revision of the NIST AI Risk Management Framework “to eliminate references to misinformation, Diversity, Equity, and Inclusion, and climate change” as well as updating federal procurement guidelines to ensure procurement of LLMs “who ensure that their systems are objective and free from top-down ideological bias.”

Encourage Open-Source and Open-Weight AI: The Action Plan calls for the federal government to “create a supportive environment for open models.” To this end, it encourages the government to ensure access to large-scale computing power for startups and academics, partner with “leading technology companies” to support the research community's access to models, data, and resources, and support adoption of open-source and open-weight models by small and medium-sized businesses.

Enable AI Adoption: To encourage adoption of AI in critical sectors that have not broadly adopted AI due to regulatory and governance risks, such as healthcare, the Action Plan advocates for developing domain-specific (e.g., healthcare, energy, agriculture) standards for AI systems and establishing regulatory sandboxes or AI Centers of Excellence to deploy and test AI tools in these contexts. The Action Plan also recommends that the Department of Defense (“DoD”) conduct intelligence assessments regarding the use of AI tools by foreign competitors and their national security establishments.

Empower American Works in the Age of AI: The Action Plan aims to support a “worker-first” AI agenda, envisioning the integration of AI literacy and skills development across the American workforce. It recommends including AI skills in education and workforce programs, clarifying tax guidance to support employer investment in AI training, and studying the impact of AI on the labor market.

Support Next-Generation Manufacturing: The Action Plan advocates for federal investment in new manufacturing technologies, enabled by AI and robotics, calling for the United States “and our trusted allies” to be manufacturers of these technologies. The Action Plan recommends using existing grant and research programs to support innovation in manufacturing and convening stakeholders to address supply chain challenges in robotics and drone manufacturing.

Invest in AI-Enabled Science: Recognizing AI’s transformative potential in scientific research, the Action Plan calls for investment in automated, cloud-enabled labs and support for organizations using AI to advance science. It encourages the public release of high-quality datasets and instituting requirements for federally-funded researchers to disclose non-sensitive datasets used in AI research.

Build World-Class Scientific Datasets: The Action Plan recognizes high-quality data as a “national strategic asset” for AI innovation and aims to make the United States a leader in AI-ready scientific data by establishing minimum data quality standards, expanding secure access to federal data, and creating secure compute environments for restricted federal data.

Advance the Science of AI: To maintain US leadership in AI research and development, the Action Plan proposes prioritizing investment in theoretical, computational, and experimental research that could yield transformative breakthroughs in AI capabilities.

Invest in AI Interpretability, Control, and Robustness Breakthroughs: The Action Plan addresses the need for greater understanding and predictability of AI systems, especially for high-stakes applications like national security. It recommends launching technology development programs focused on interpretability, control, and robustness, prioritizing these areas in federal research, and coordinating a hackathon to test AI systems.

Build an AI Evaluations Ecosystem: To ensure reliable and trustworthy AI, the Action Plan supports the development of rigorous evaluation methods for AI systems. The Action Plan also acknowledges that evaluations could aid regulators in their application of existing law to AI systems. It calls for the publication of federal guidelines for AI evaluation, supporting the development of science for measuring AI models, and developing AI testbeds for piloting AI systems in secure, real-world settings.

Accelerate AI Adoption in Government: The Action Plan notes the potential for AI to improve the delivery of government services. It seeks to transform government operations through AI by formalizing interagency coordination on AI and creating talent exchange programs to rapidly detail AI talent to agencies in need. The Action Plan also recommends the development of a procurement toolbox so that agencies can easily choose among multiple models in a manner compliant with relevant privacy, data governance, and transparency laws.

Drive Adoption of AI within the Department of Defense: The Action Plan supports aggressive adoption of AI by the DoD. It recommends that the DoD assess the AI skills and talent needed across its workforce and implement targeted development programs to meet these requirements. In addition, the Action Plan calls for the creation of an “AI & Autonomous Systems Virtual Proving Ground,” and for the development of streamlined processes to identify workflows that could be automated with AI. The Action Plan also prioritizes securing priority access to advanced computing resources for the DoD in national emergencies.

Protect Commercial and Government AI Innovations: The Action Plan calls for close collaboration between government and industry to safeguard AI innovations from security threats, including cyberattacks and insider risks, ensuring that American AI leadership is not undermined by malicious actors.

Combat Synthetic Media in the Legal System: The Action Plan recommends the development of forensic guidelines and benchmarks for deepfake detection, and exploring additions to the Federal Rules of Evidence related to deepfakes.

Pillar 2: Build American AI Infrastructure

The Action Plan’s second pillar is aimed at developing infrastructure necessary to support AI, with a particular focus on energy generation.

Create Streamlined Permitting for Data Centers, Semiconductor Manufacturing Facilities, and Energy Infrastructure while Guaranteeing Security: To accelerate the construction of critical AI infrastructure, including semiconductor factories, data centers, and power sources, the Action Plan proposes reforming environmental permitting processes and making federal lands available for development, while ensuring that infrastructure does not rely on technology, including information and communications technology and services (“ICTS”), provided by foreign adversaries.

Develop a Grid to Match the Pace of AI Innovation: Recognizing the increasing energy demands of AI, the Action Plan recommends optimizing the current electric grid using advanced technologies, preventing premature decommissioning of power resources, and prioritizing the interconnection of advanced energy sources, such as nuclear and geothermal, to ensure reliable and affordable power.

Restore American Semiconductor Manufacturing: The Action Plan aims to “bring semiconductor manufacturing back to US soil” by eliminating “extraneous policy requirements” from CHIPS Act projects, streamlining regulations related to semiconductor manufacturing efforts, and ensuring that grant programs focus on accelerating the integration of AI tools into chip production.

Build High-Security Data Centers for Military and Intelligence Community Usage: Noting the potential for AI systems to process highly sensitive data for intelligence and military applications, the Action Plan recommends the creation of new technical standards for high-security data centers and the adoption of classified compute environments to support scalable and secure AI workloads.

Train a Skilled Workforce for AI Infrastructure: Highlighting the need to develop a workforce capable of building and maintaining AI infrastructure, the Action Plan recommends identifying priority occupations in AI infrastructure, supporting the creation of industry-driven training programs, and expanding early career apprenticeship opportunities.

Bolster Critical Infrastructure Cybersecurity: The Action Plan emphasizes the need to strengthen the cybersecurity of America’s critical infrastructure in the face of rapidly advancing AI capabilities. It notes that while AI systems themselves can be excellent defensive tools for critical infrastructure operators, the use of AI in critical infrastructure can expose those systems to adversarial threats. The Action Plan proposes the establishment of an AI Information Sharing and Analysis Center (“AI-ISAC”), led by the Department of Homeland Security, to facilitate the sharing of AI-specific threat intelligence and best practices across critical infrastructure sectors. It also recommends the development of guidance for the private sector on remediating AI-specific vulnerabilities.

Promote Secure-By-Design AI Technology and Applications: The Action Plan calls for updates to existing AI standards and frameworks within the DoD and the intelligence community to strengthen AI assurance.

Promote Mature Federal Capacity for AI Incident Response: The Action Plan acknowledges that the proliferation of AI technologies necessitates prudent planning to ensure that in the event of a system failure, impacts to critical infrastructure and services are minimized. It recommends the inclusion of AI in incident response frameworks and standards, and the modification of CISA’s playbooks and incident response plan to address considerations for AI systems.

Pillar 3: Lead in International AI Diplomacy and Security

The last pillar outlines steps intended to “drive adoption of American AI systems, computing hardware, and standards throughout the world . . . while preventing [US] adversaries from free-riding on our innovation and investment.”

Export American AI to Allies and Partners: The Action Plan seeks to create an AI alliance of countries using a US-developed AI technology stack, including hardware, models, software, applications, and standards. Such a proposal intends to prevent countries from turning to rivals for technology.

Counter Chinese Influence in International Governance Bodies: The Action Plan states that international efforts to date “have advocated burdensome regulations, vague ‘codes of conduct’ that promote cultural agendas that do not align with American values, or have been influenced by Chinese companies.” To counter these identified issues, the Action Plan recommends that the government advocate for international AI governance approaches that promote innovation, reflect American values, and counter authoritarian influence.

Strengthen AI Compute Export Control Enforcement: While the Action Plan emphasizes the importance of export promotion for the United States AI technology stack, it reflects a continued emphasis on the importance of export controls to “deny our adversaries access to [advanced AI compute]” as a matter of “both geostrategic competition and national security.”  In so doing, it makes certain high-level recommendations, including calling on the Department of Commerce (“Commerce”) and other agencies to (i) collaborate with industry to explore leveraging location verification features on advanced AI compute to ensure it does not end up in countries of concern, and (ii) collaborate with the intelligence community to monitor emerging technology developments in AI compute to ensure full coverage of possible countries or regions in which chips are being illicitly diverted, which could then support end-use monitoring.

Plug Loopholes in Existing Semiconductor Manufacturing Export Controls: The Action Plan also calls for greater measures to prevent foreign adversaries from using US semiconductor technologies. It proposes the development of new export controls on sub-systems of semiconductor manufacturing equipment.

Align Protection Measures Globally: The Action Plan emphasizes the need for strong export controls on sensitive AI technologies and urges US allies to adopt similar measures, discouraging backfilling by foreign partners. It calls for developing and sharing complementary technology protection strategies, including in research and higher education, to mitigate risks from adversaries and concerning entities. The Action Plan also contemplates ensuring complementary technology protection measures by incentivizing allies through application of “secondary tariffs” or export control regulatory tools, such as the Foreign Direct Product Rule, while also promoting plurilateral controls, outside of multilateral export control arrangements, across the AI technology stack. It further identifies a need to work with allies to prohibit US adversaries from supplying their defense-industrial base or acquiring controlling stakes in defense suppliers.

Ensure that the US Government is at the Forefront of Evaluating National Security Risks in Frontier Models: The Action Plan prioritizes the evaluation of advanced AI systems for national security risks, including cyber and CBRNE threats, and calls for recruiting top AI researchers to maintain cutting-edge federal evaluation capabilities. It also proposes the evaluation of AI systems provided by adversaries for backdoors and malicious code.

Invest in Biosecurity: To address the risk that malicious actors will use AI to create harmful pathogens and other biomolecules, the Action Plan proposes robust screening and verification of end users, working with allies to ensure international adoption of effective screening tools. Specifically for nucleic acid synthesis tools, the Action Plan calls for mechanisms to facilitate data sharing between nucleic acid synthesis providers to detect malicious actors, and for the creation of mechanisms beyond voluntary attestation to ensure robust verification.

Executive Order on Promoting the Export of the American AI Technology Stack

This Executive Order directs the Department of Commerce to establish the “American AI Exports Program,” a coordinated initiative to advance the full stack of US-origin AI technologies globally, within 90 days. The purpose of the program is to ensure that American-developed AI technologies, standards, and governance frameworks become the foundation for international adoption, thereby deepening strategic ties with allied nations, reinforcing the United States’ position as a global leader in AI innovation, and addressing the need to “decrease international dependence on AI technologies developed by [US] adversaries.”  Consistent with the Action Plan, it also tasks Commerce, the State Department and other agencies with taking steps to facilitate the export of American AI technology stacks to allies and partners.

This Executive Order lays out a multi-pronged approach for the program. First, Commerce will solicit proposals from “industry-led” consortia on a “full-stack AI technology package” to prioritize for the program, including information on their full-stack AI technology package, business model, security compliance measures, and target markets. Second, once Commerce selects priority AI export packages, federal entities are directed to coordinate deployment of federal financing tools to support AI export packages. The State Department also is tasked with developing and executing a strategy to promote the export of AI technologies and standards as well as coordinate US participation in multilateral and unilateral efforts for AI deployment and export promotion.  

Executive Order on Accelerating Federal Permitting of Data Center Infrastructure

This Executive Order directs various federal agencies to take steps to ease regulatory burdens to support buildout of AI infrastructure, including data centers, semiconductor manufacturing, and energy projects. This Executive Order implements the strategic objective outlined in the Action Plan to simplify permitting processes for essential AI infrastructure—including data centers, semiconductor manufacturing, and energy projects.

The primary provisions of the Executive Order apply to "Qualifying Projects," which include both "Data Center Projects" and "Covered Component Projects" that meet specified criteria—including exceeding a defined capital expenditure or electrical load increase, contributing to national security, or being specifically designated by the government as Qualifying Projects.:

• A “Data Center Project” is a facility that requires greater than 100 megawatts (“MW”) of new load dedicated to AI inference, training, simulation, or synthetic data generation.
• A “Covered Component Project” is infrastructure comprising Covered Components, or a facility with the primary purposes of manufacturing or otherwise producing Covered Components.
• “Covered Components” include energy systems and equipment, semiconductors and associated materials, networking equipment, and data storage solutions, including hardware, software, and services.

This Executive Order directs federal agencies to streamline environmental reviews, permitting, and land access for Data Center and Covered Component Projects. In addition, it orders Commerce to launch an initiative to provide financial support for these projects, including loans, grants, tax incentives, and offtake agreements. This Executive Order also revokes EO 14141, Advancing United States Leadership in Artificial Intelligence Infrastructure, issued by President Joe Biden.

Executive Order on Preventing Woke AI in Federal Use

This Executive Order asserts that “diversity, equity, and inclusion” poses an existential threat to reliable AI and highlights examples of AI models “prioritizing DEI requirements at the cost of accuracy.” This Executive Order directs agency heads to only procure LLMs that are developed in accordance with “truth-seeking” and “ideological neutrality” principles (referred to as Unbiased AI Principles), meaning that the LLM is truthful in its responses, prioritizes accuracy and objectivity, and is not intentionally encoded with partisan or ideological judgements.

To support this direction, the Office of Management and Budget (“OMB”) is tasked with issuing guidance to agencies on procuring LLMs developed in accordance with the Unbiased AI Principles. This includes requiring contract terms for LLMs specifying compliance with the Principles and providing that decommissioning costs will be charged to the vendor in the event of termination by the agency for the vendor’s noncompliance. It also notes that the guidance should “avoid over-prescription” and afford latitude for vendors to comply with the Unbiased AI Principles. Notably, the EO directs OMB to make exceptions “as appropriate” for the use of LLMs in national security systems.