Overview
On July 23, 2025, the White House released Winning the Race: America's AI Action Plan, a sweeping roadmap outlining the administration's strategy to cement US dominance in artificial intelligence (AI). Here, we provide a high-level overview of the Action Plan's three pillars, along with some key takeaways spanning several substantive areas. Future client updates will provide more detailed analyses of the plan, grouped by subject area. The main takeaway is that, regardless of your industry, having a smart action plan for integrating AI into your business model and government affairs strategy will yield significant dividends. Under the Plan, your company’s ability to advance the United States' competitiveness in AI will be a skeleton key for unlocking federal government support and favorable regulatory reform.
Created under Executive Order 14179, the Plan promotes a deregulatory, innovation-driven agenda focused on private-sector leadership, reduced red tape, and expanded global influence. Framing AI as a national security and economic imperative, the Plan declares that the "United States needs to innovate faster and more comprehensively than our competitors in the development and distribution of new AI technology across every field, and dismantle unnecessary regulatory barriers that hinder the private sector in doing so."
The Action Plan's Three Pillars
The AI Action Plan is organized around three pillars: accelerating AI innovation, building American AI infrastructure, and leading in international AI diplomacy and security.
I. Accelerating AI Innovation
Framing regulation as a key roadblock to progress, Pillar I directs the federal government to clear the way for private-sector-driven innovation in AI. The innovation pillar of the AI Action Plan calls for repealing or revising regulations deemed burdensome, including holdovers from President Biden's previously rescinded AI-focused Executive Order 14110. Notably, Pillar I includes a similar, arguably vaguer, provision than the AI law moratorium that the Senate stripped from the One Big Beautiful Bill Act. The AI Action Plan provides that the Federal government "should not allow AI-related Federal funding to be directed toward states with burdensome AI regulations that waste these funds," while also "not interfer[ing] with states' rights to pass prudent laws that are not unduly restrictive to innovation." It directs the Office of Management and Budget (OMB) to work with Federal agencies that have AI-related discretionary funding programs to ensure that they "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."”
Pillar I, Accelerate AI Innovation, is of particular relevance to those in regulated industries, because it outlines a pro-innovation agenda with potentially far-reaching consequences, such as in the financial sector for investment banks, investment advisers, public companies, and other capital markets participants.
Pillar I reflects a deliberate policy shift toward minimizing regulatory friction in the development and deployment of AI technologies. The Plan directs OMB, in coordination with all federal agencies, to identify and repeal regulations that may "unnecessarily hinder" AI innovation.
For example, this approach signals a potential reorientation of regulatory posture at the Securities and Exchange Commission (SEC) and the Federal Trade Commission (FTC). The Plan explicitly calls for a review of FTC investigations and consent decrees from the prior administration, suggesting a rollback of enforcement theories that may have previously constrained AI experimentation in regulated industries.
Notably, the Plan encourages the creation of regulatory sandboxes and AI Centers of Excellence, including in collaboration with the SEC. These initiatives are designed to facilitate the rapid deployment and testing of AI tools in a controlled environment. For many industry participants, these initiatives present a unique opportunity to pilot AI applications without the immediate burden of full regulatory exposure, such as in generative AI for research, machine learning for product, portfolio, or customer service optimization, or natural language processing for compliance monitoring. We expect that participation in such programs will require transparency around model design, data provenance, and risk mitigation strategies.
The Plan also emphasizes the development of a national AI evaluations ecosystem, led by the National Institute of Standards and Technology (NIST). This initiative aims to establish standardized methodologies for assessing AI system performance, reliability, and compliance with existing law. For public companies and other businesses in regulated industries, this development may foreshadow heightened expectations around internal controls, model validation, and disclosure practices, particularly where AI is material to business operations or reporting. We expect that existing disclosure, oversight, and minimum-standard frameworks will evolve to incorporate these emerging standards.
The AI Plan also calls for the expansion of open-source and open-weight AI models, with the federal government playing a facilitative role in ensuring access to compute resources and fostering a robust financial market for AI infrastructure. Although this approach may lower barriers to entry for smaller firms and innovators to enter the data center and capacity space, it also raises questions about model governance, intellectual property, and systemic risk — regulators may be compelled to address these issues in future rulemakings or guidance.
Finally, the Plan includes provisions to promote AI literacy and workforce development, including tax incentives for employer-sponsored training programs. Companies may wish to explore these incentives in their current implementation phase before they are finalized as part of broader talent strategies, particularly as AI reshapes roles in research, operations, and client engagement.
Pillar I of the AI Action Plan articulates an innovation-forward vision for AI in the US economy. For companies in regulated industries, it presents both opportunities and challenges: the prospect of accelerated innovation and regulatory flexibility, tempered by the need for robust governance, transparency, and strategic foresight. Notably, the innovation pillar of the plan does not directly address intellectual property, including issues relating to patents or copyrights, despite these issues being the subject of considerable policy debate and discussion. Firms should begin assessing their AI strategies in light of these developments and consider proactive engagement with regulators and industry consortia shaping the future of AI oversight.
II. Building American AI Infrastructure
Pillar II of the Action Plan focuses on rapidly expanding and stabilizing the infrastructure and energy systems needed to power large-scale AI deployment, emphasizing sweeping reforms to streamline permitting processes, modernizing the electric grid and accelerating the construction of data centers, semiconductor manufacturing facilities, and energy infrastructure. The plan calls for easing environmental reviews, expanding the use of federal land for data centers, and prioritizing dispatchable energy sources, like nuclear and geothermal, while making workforce investments to train the personnel required to build and maintain these systems. The administration also emphasizes stabilizing the current grid by safeguarding existing power generation assets, preventing premature decommissioning, and securing compliance with nationwide standards for resource adequacy. This focus includes identifying means to leverage backup power sources to enhance grid reliability during peak demand and ensuring an uninterrupted and affordable supply of power.
To optimize existing resources, the initiative promotes strategies like advanced grid management technologies and upgrading power lines to increase transmission efficiency along current routes. It also includes exploring new methods for large power consumers, such as data centers, to manage consumption during critical periods to boost grid reliability and unlock capacity. Prioritizing the integration of reliable, dispatchable energy sources — such as enhanced geothermal, nuclear fission, and fusion — is central to the plan, alongside reforming power markets to align financial incentives with grid stability needs.
The Plan (in Pillar I) also encourages exploration of the use of financial markets, such as spot and forward markets for commodities to accomplish these goals. In the spot markets, buyers and sellers agree on a price, and trade almost immediately so resources can be used. The forward market involves agreements to buy or sell at a set price on a future date, providing the ability to budget. While forward markets are often used for hedging against price fluctuations, spot markets are highly liquid and used for immediate transactions. Financial products like futures, forwards, options, and swaps can be used to manage risks. Finally, the administration envisions creating a strategic blueprint to guide the US energy system through the challenges of the 21st-century landscape. By stabilizing today’s grid, optimizing existing resources, and expanding infrastructure for the future, the nation aims to meet the energy demands of large-scale AI deployment while ensuring a reliable and affordable power grid for all Americans.
III. Leading in International AI Diplomacy and Security
Pillar III outlines a strategy to establish the US as the global standard-setter for AI, while countering adversarial influence, particularly from China. This section emphasizes promoting American-developed AI systems, chips, and technical standards abroad, and supporting allies in adopting US infrastructure and governance approaches. In this regard, the President issued an Executive Order on June 23, 2025 establishing that, within 90 days, the Secretary of Commerce, in consultation with the Secretary of State and the Director of the Office of Science and Technology Policy (OSTP), will implement the American AI Exports Program (Program) to support the development and deployment of United States full-stack AI export packages, including hardware (e.g., chips, servers, and accelerators), software (e.g., AI models and security/cybersecurity systems), technology, including data pipelines and labeling systems, and services (e.g., data center storage, cloud, and networking services). Furthermore, this Executive Order establishes that the foregoing agencies and the Secretaries of Defense and Commerce shall evaluate submitted proposals for inclusion under the Program, as well as to coordinate mobilization of federal financing tools in support of priority AI export packages.
Pillar III also seeks to tighten enforcement of export controls on advanced AI export packages, including through: (1) recommending location verification and end-use monitoring on advanced AI computing chips (i.e., integrated circuits); (2) implementing new export controls on semiconductor manufacturing sub-systems and components; and (3) imposing additional export controls, including under the Foreign Direct Product Rule and other secondary sanctions restrictions, on third countries that do not align their restrictions with US rules or otherwise backfill for prohibited destinations, end users, and end uses. This section emphasizes the administration’s national security priorities, proposing new standards for evaluating frontier AI systems for national security risks and countering Chinese influence in international standards governing bodies.
Key Takeaways
The AI Action Plan offers a clear signal that federal AI policy will focus on accelerating private-sector development by reducing regulatory burdens and aggressively pursuing global dominance. Companies working in AI, critical infrastructure, energy, national security, and defense should monitor implementation closely, particularly in areas related to procurement, permitting, and supply chain security.
Often, a major bottleneck in the ability to develop and bring data centers online is the ability to secure and develop the necessary energy and infrastructure to support the addition of such large energy loads. Pillar II of the AI Action Plan outlines ambitious reforms and initiatives to streamline this process and incentivize market reform and innovation to enhance the process and allow more speed to market. Although the AI Action Plan offers broad directives and goals, the way in which these directives and goals will be implemented across the varying regions and markets will determine the success of the Plan. Much of the Plan’s success will likely be dictated by the federal agencies, but state agencies and regional market operators will play a large role in the Plan’s success as well. The Plan's success also hinges on how these reforms impact financing opportunities and how the costs of the necessary infrastructure to support data centers will be allocated among the various customer groups using the nation’s energy grid.
At the same time, the Plan leaves several major questions unanswered. It does not address how the government will regulate agentic AI, nor does it offer a framework for handling intellectual property or privacy concerns. Notably, the Plan does not mention applicable privacy protections in the United States, such as under the Privacy Act or Health Insurance Portability and Accountability Act (HIPAA), which would need to be considered in a variety of contexts when dealing with data sets that include personally identifiable information. These omissions raise uncertainty for developers, platforms, content owners, and service providers that are increasingly navigating legal risk in these areas.
Our interdisciplinary AI team has collated the following key takeaways as related to specific aspects of the Action Plan or specific sectors:
- Use of AI by the Federal Government in Enforcement: The federal government, particularly the SEC and the Department of Justice, have long-used AI to identify conduct and further ongoing investigations into fraud, market manipulation, and even foreign bribery cases. Since portions of the Plan describe sharing AI resources and expertise across agencies, there is the potential for these existing use cases to be expanded and aligned with the administration's key enforcement goals, such as combatting waste, fraud, and abuse, tariff evasion, and other trade-related cases.
- Conflicts between State and Federal Requirements: The Plan is the first step in a deregulatory agenda for AI that will likely be implemented in detail by various federal agencies and officials in the coming months. As a result, state requirements may come into conflict with federal efforts, and organizations may be placed in the unenviable position of needing to weigh state compliance with potential blowback from federal officials, particularly in the absence of clear federal guidance or law. In addition, states may more aggressively pursue regulations to counter federal efforts and pursue claims against the administration, such as challenging deprivation of discretionary AI funds as imposing unconstitutional conditions. Understanding the interplay between state and federal requirements and the nuances of preemption will be crucial to successfully navigating potential conflicts. Our team of former state and federal government officials can help with questions in this area.
- Consumer Protection: The administration has signaled a desire to enforce consumer protection statutes in some instances, such as when AI is used to deceive customers or misrepresent the capabilities of products. A recent example was a case in which the allegedly misleading use of a "robot lawyer" led to a fine from the FTC. At the same time, the administration has signaled a desire to roll back certain consumer protection enforcement efforts that might, in its view, conflict with certain First Amendment principles.
- Data Centers and Energy: The administration has emphasized the importance of high-speed data centers for a competitive AI future. Accordingly, the AI Plan has set forth numerous goals that will roll back regulations that might encumber the permitting or establishment of those data centers, while also ensuring that the computing power in those data centers is enhanced by American semiconductors. To the extent that more energy is needed to support data centers, the action plan also seeks to limit regulations, including environmental regulations, which might have been a roadblock to the establishment of data centers in the past.
- Eradicating Misinformation, DEI Terms, and Climate Change Terms: The administration has viewed some social media companies and various industries as engaging in a partisan approach in the development and operation of algorithms and implementation of user policies. Further, misinformation, particularly in the election context, has been an area affecting national security where nation states attempt to push alleged misinformation in elections. The Action Plan seeks to eliminate references to misinformation, diversity, equity, and inclusion terms, and climate change, ensuring that the federal government only contracts with LLMs that have no ideological bias. It also calls to research whether frontier models from China align with Chinese Communist Party talking points and censorship.
- Capital Markets: For capital markets participants, such as broker-dealers and investment advisers, the AI Plan presents both opportunities and challenges: the prospect of accelerated innovation and regulatory flexibility, tempered by the need for robust governance, transparency, and strategic foresight. Firms should begin assessing their AI strategies in light of these developments and consider proactive engagement with regulators and industry consortia shaping the future of AI oversight.
- Risk of Fraud in AI Space: Although the administration plans to take a "hands off" approach to the regulation of AI, investors should also beware of bad actors who seek to profit from deregulation. We expect to see large amounts of capital raised for AI by numerous types of entities in different industries. Investors and deal makers should be cautious and conduct thorough due diligence before investing in or merging with an AI entity, particularly one that is new to the marketplace. This approach is especially warranted if certain enforcement or investigation efforts by federal officials come to be viewed as unnecessarily interfering with the administration's goal of US dominance in AI.
- Government Data and Procurement: The Plan includes a number of initiatives to make government computing power and data available to industry through national laboratories and other organizations. It also emphasizes the importance of making both industry and government data sets widely available. However, the Plan leaves open the possibility that access to numerous government data sets may be controlled through restricted platforms and compute environments that only permit access by projects that are deemed valuable to the government. This approach follows a recent trend in which the government has been leveraging its data as an asset like the commercial industry instead of a public resource. In addition, the Plan suggests that industry data should be made available when deemed "non-proprietary," but there can be differences of opinion about whether public or third-party data that a company spends resources collecting and curating falls within this category. Our team has extensive experience negotiating data sharing agreements with government agencies and laboratories and will be tracking developments under such agreements and government procurement more generally as the Plan is implemented.
- Intellectual Property: The Action Plan says surprisingly little about intellectual property, despite important issues around copyrights, patents, and trade secrets taking center stage in policy debates. One will expect that the administration will maintain a deregulatory posture in these areas as well, particularly when it comes to potential requirements around the licensing of AI training data. We can expect to see increased emphasis on the use of AI to innovate both within the Copyright Office and the Patent and Trademark Office. Tellingly, on the same day that the AI Plan was released, President Trump suggested that it is not possible to have a successful AI program if licensing payments were required for AI training data. Still, this is an area where the Courts and Congress, rather than the administration, are likely to take center stage.
- National Security and Export Controls: Consistent with prior Trump administration statements and policies, the Action Plan emphasizes the need for enhancement of national security regulations, particularly export controls, and stricter enforcement of existing rules. It also makes clear that the US intends to be active in the international arena by seeking to counter the influence of China in various international bodies and pushing third countries to adopt similarly robust national security controls.
- Litigation: The Action Plan's emphasis on using AI across the workforce, including in the legal sector, underscores the continued need for lawyers to become aware of AI use cases in legal practice. Moreover, the Action Plan’s recognition of the need for the Federal Rules of Evidence to adapt to AI-based evidence is a welcome development and aligns with recent efforts to do the same. Litigants are already introducing AI-based evidence in Court proceedings. Courts must stay informed and adjust.
- Privacy: The Action Plan says very little about privacy-related issues, underscoring that the administration does not intend to promote regulation or policy in this area. Therefore, stakeholders will need to continue to exercise their familiarity with current federal and state policy law and understand how those laws will impact AI-related data, despite the deregulatory posture of the Agency.
We will continue tracking regulatory developments across federal agencies and legal developments at the state level through Steptoe's AI Legislative Tracker. Please contact members of our interdisciplinary artificial intelligence team spanning our Commercial, Consumer and Government Litigation, Energy, Financial Innovation and Regulation, Intellectual Property, International Trade and Regulatory Compliance, Investigations and White Collar Defense, and Telecom and Technology practices to assess how the AI Action Plan may affect your operations or investments. In addition, our Government Affairs and Public Policy team is ready to assist companies in working with agencies and offices responsible for implementing the AI Action Plan by leveraging both our experience in engaging with the administration and Congress, as well as our insight into industry-specific impacts of AI deregulation.
