Trump’s ‘America First’ AI Policy

Summary

President Trump’s AI policy positions AI as a crusade for global dominance, achieved through the politicisation of technology and aggressive deregulation.  The approach is fraught with long-term risks, including widening societal discrimination domestically.

In his second term, President Donald Trump has comprehensively restructured the US federal government’s role in regulating the development of artificial intelligence (AI). The Trump administration passed Executive Order (EO) 14179, ‘Removing Barriers to American Leadership in Artificial Intelligence’, in January 2025.[1]  The EO framed AI advancement as a zero-sum geopolitical ‘race’ crucial to maintaining America’s hegemony, and pitched it as equal to the space race; the US must win at all costs.  It advocates a ‘whole of government’ approach to dismantling obstacles to private-sector innovation. This restructuring is justified by threats on multiple fronts, including intensified zero-sum competition with China. The Trump administration also released ‘Winning the Race: America’s AI Action Plan’ in July 2025. Key aspects of the Action Plan are enumerated below.

First Pillar: Accelerating Innovation

The AI Action Plan initiative aims to eliminate rules that impede innovation in AI. It seeks to prevent fraudulent, deceptive, or unfair business practices and anti-competitive mergers that could lead to monopolies.[2] The AI Action plan directs federal agencies to consider a state’s AI regulatory climate when awarding AI-related discretionary funds. This means that states that implement the administration’s ‘burdensome’ AI rules (e.g., on safety or bias) risk being denied federal funds. The goal is to financially pressure states into adopting the federal government’s pro-innovation, deregulatory approach rather than creating a ‘patchwork’ of different rules. This strategy aims to create a uniform and minimally regulated framework for AI development.[3]

The plan’s deregulation efforts are designed to actively promote the open distribution of AI models, primarily ‘open-source’ and ‘open-weight’ systems developed with federal money. Its primary goals include democratising AI by giving small- and medium-sized businesses access to the same advanced tools as large tech companies, and establishing American technology and values as the global standard by encouraging worldwide adoption of these open, US-backed systems. The plan also directly addresses the impact of AI automation on employment. It establishes a worker-first AI agenda that involves creating and funding new AI employment research centres. These centres will analyse how automation is specifically affecting the job market and use that research to develop and federally fund new training and retention strategies, helping the labour force adapt to an increasingly automated economy.[4]

Second Pillar: Building AI Infrastructure

The second pillar of the action plan is built on the realisation that software and hardware alone do not guarantee AI dominance, it also requires a robust physical infrastructure and energy to support it. The main feature of this strategy is its focus on constructing data centres, semiconductor manufacturing facilities and energy infrastructure to support them. The plan also calls for oiling the mechanisms to speed up reviews for environmental clearance under foundational laws like the Clean Air Act, the Clean Water Act and the National Environmental Policy Act (NEPA). It even proposes creating a new NIPA category to exclude data centre-related actions, and a nationwide Clean Water Act permit that would waive specific requirements, thereby reducing construction timelines.[5]

The plan recognises that AI is an energy-intensive industry. Thus, it calls for a strategy to ‘develop a grid to match the pace of AI innovation’.[6] This includes two simultaneous efforts: improving the existing grid by addressing the ‘premature decommissioning’ of legacy power plants and maximising existing resources, as well as incorporating ‘frontier’ energy resources such as nuclear fission, fusion and geothermal energy.  This energy policy is seen as a national security imperative, highlighting that the ‘American energy capacity has remained stagnant since the 1970s while China rapidly built out its grid’.[7]

The infrastructure pillar also acknowledges that physical construction would require a skilled workforce.  Hence, the plan calls for actions to increase the number of such technicians through active engagement with states and local governments, as well as through industry-driven training programmes, including expanding the base of registered apprenticeships.[8]

Third Pillar: Leading in International Diplomacy and Security

The final pillar of the action plan presents the ‘America First’ foreign policy agenda for the AI era. It moves away from the standard multilateral setting towards a more transactional, unilateral approach to project American power globally and secure its technological dominance.[9] The central aspect of this strategy is to export ‘the American AI technology stack’. The plan mandates the U.S. Department of State (the federal agency for foreign policy) and the U.S. Department of Commerce (the federal agency for trade) to collaborate with private US industries to deliver secure, full-stack AI packages, including models, software, hardware, applications and standards, to partner nations. This aims to create an ‘enduring global alliance’ around American technology, preventing any American adversary from making its allies dependent on the United States.[10]

This proactive export strategy is coupled with a defence posture to counter China’s advancement. The plan explicitly mentions efforts to counter Chinese influence in setting norms and international governance bodies. It also calls for mending loopholes in export controls on critical technologies, especially semiconductor manufacturing equipment, to advance semiconductor manufacturing and restrict rivals’ access to the crucial components of AI.[11]

A Unilateral Ideologically Charged Approach

America’s AI Action Plan is an ambitious doctrine for securing technological dominance for the US. However, its unilateral, aggressive and ideologically charged approach is fraught with internal contradictions and several significant risks that can undermine its objectives and affect the US’s foreign relations and domestic policy, as well as the ethical development of AI.

The plan’s core principle, which adopts a zero-sum AI race with China, reflects a withdrawal from multilateral cooperation to a unilateral stance, amplifying strategic competition. The Trump administration’s approach will inevitably harm crucial partners, such as the European Union, which prioritises a rights-based approach to AI governance through its AI Act 2024. The EU prioritises protecting fundamental human rights above all other considerations, such as innovation, profit, or geopolitical competition, unlike the US’s deregulatory and unilateralist stance. This divergence creates challenges for transatlantic commerce, restrictions on data circulation, and impediments to collaborative ventures in AI governance and standards-setting.

Moreover, many have argued that America’s approach of exporting ‘the American AI stack’[12] is a new form of technological neocolonialism. The ‘all or nothing sales pitch’ creates an overdependence on US technology and corporate houses, giving China an undue advantage and making it an attractive alternative for countries seeking to maintain their technological sovereignty.[13]

Trump’s AI plan accelerates the construction of data centres, semiconductor facilities and power infrastructure by weakening environmental reviews and waiving long-standing protections under laws such as the National Environmental Policy Act (NEPA) and the Clean Water Act. This benefits private companies by lowering costs and speeding deployment, but shifts the environmental and social burdens onto local communities through increased water extraction, higher energy demand, and reduced oversight of resource use. In this sense, a policy that eases participation for AI developers can simultaneously erode equity and sustainable development in the regions hosting the physical backbone of the AI industry.

Although Trump’s AI action plan depends on a rapid scaling of domestic semiconductor manufacturing, it sits uneasily alongside the administration’s broader protectionist trade agenda. Semiconductor production is one of the most globally interdependent industries, relying on foreign suppliers for critical chemicals, lithography equipment, speciality gases, wafers and packaging. Trump’s emphasis on tariffs and trade confrontations—with China but also periodically with allies such as Japan, South Korea and the EU—risks disrupting these complex supply chains. As a result, the very trade volatility created by ‘America First’ policies could undermine the stable flow of components needed to build the semiconductor base on which the AI strategy relies, creating an internal contradiction at the heart of the plan.

Perhaps, the most controversial aspect of Trump’s AI doctrine is its ideological battle against what it calls ‘woke AI’.  This is codified through the executive order ‘Preventing Woke AI in the Federal Government’ issued in July 2025,[14] which directs federal agencies to take only those large language models (LLMs) that are flagged ‘free from ideological biases’ and designed to pursue ‘objective, truth, rather than social engineering agendas’.[15] This aspect also represents the politicisation of technology. The government can leverage its immense procurement power to force private technology companies to alter their products to conform to the administration’s political view.

The National Institute of Standards and Technology (NIST), a neutral technical body, has been directed to revise its AI risk management framework to remove all references to ‘misinformation, diversity, equity and inclusion and climate change’.[16] This directive was a key component of the Trump administration’s AI policy, unveiled on 23 July 2025.

The executive order on Woke AI functions as more than just a statement of political preference. It reshapes the market mechanism by reengineering the technology sector by linking lucrative federal contracts, including the AI budgets of the Department of Defense, to comply with vague and politically charged standards of ‘truth seeking’ and ‘ideological neutrality’. This procurement process incentivises technology firms to alter the products they offer to align with the administration’s political agenda.[17]

The administration’s policy presents a contradictory dynamic with the scientific community, notably the National Science Foundation (NSF). While it charges the NSF with significant responsibilities, such as developing AI research labs and expanding access to computing power, it simultaneously undermines the agency. Reports indicate that the administration has cancelled over 1,600 grants for political reasons, dismissed staff from NSF’s division focused on equity in STEM, and replaced peer-review processes with ideological assessments. This creates a paradox in which the NSF is expected to carry out vital tasks while facing the erosion of its resources and integrity.[18] By plaguing public institutions like the NSF, the US might rely entirely on a handful of private industrial labs for frontier AI development, losing the diverse, publicly funded research ways that offered a crucial counterbalance to concentrated corporate power and ensured that technological advancement helps the larger public interest.[19]

The plan’s aggressive, defence-first focus has resulted in significant inter-agency conflicts, particularly over funding and authority for advanced technologies such as biotechnology and quantum technologies. For instance, the policy’s emphasis on the Department of Defense leading applied research has sparked funding and authority disputes with the Department of Energy (DoE) and the NSF, which traditionally manage fundamental R&D in these areas.[20] This has also created policy contradictions: the Department of Commerce is tasked with promoting AI exports, while the DoD simultaneously seeks to restrict those same dual-use technologies, citing national security risks.[21]

Moreover, state-level measures directly contradict the Trump administration’s federal vision for AI, which seeks to eliminate what it calls burdensome safety and consumer-protection rules and to establish a uniform, minimally regulated national market. By contrast, laws such as New York’s ‘Artificial Intelligence Companion Models’ introduce precisely the kinds of mandatory safety features, disclosures and oversight mechanisms that the federal plan intends to pre-empt. The proliferation of such state regulations fragments the national regulatory landscape, undermines the administration’s deregulatory agenda, and reveals a widening gap between federal priorities and state-level public-interest concerns.

At the same time, a key component of ‘America’s AI Action Plan’ is gaining traction. This refers to the plan’s strategy to promote the ‘export of the American AI technology stack’—a policy encouraging industry-led coalitions to bundle and sell integrated packages of US hardware (like advanced chips), cloud services and AI models to allied nations.[22] The US–Japan Technology Prosperity Deal, signed in October 2025, is the first significant example of this policy in action. This bilateral agreement establishes a ‘pro-innovation framework’ that prioritises rapid, market-led AI development by aligning US and Japanese regulatory standards.[23] By strengthening technological partnerships and supply chain resilience with key allies like Japan and South Korea, the US aims to create a unified democratic bloc to serve as a direct market and governance alternative to China’s ‘Digital Silk Road’ and its model of state-controlled AI.[24]

To conclude, the AI policy under President Trump marks a radical transformation to an aggressive America First doctrine. This positions AI as a crusade for global dominance, achieved through the politicisation of technology and aggressive deregulation.  The approach is fraught with long-term risks, including widening societal discrimination domestically.

Views expressed are of the author and do not necessarily reflect the views of the Manohar Parrikar IDSA or of the Government of India.

[1] “Removing Barriers to American Leadership in Artificial Intelligence”, The White House, 23 January 2025.

[2] “Artificial Intelligence”, Federal Trade Commission, 2025.

[3] Holly Fechner et al., “Trump Administration Issues AI Action Plan and Series of AI Executive Orders”, Covington, 25 July 2025.

[4] “Trump Administration Releases AI Action Plan and Three Executive Orders on AI: What Employment Practitioners Need to Know”, Seyfarth, 25 July 2025.

[5] Stuart D. Levi et al., “White House Releases AI Action Plan: Key Legal and Strategic Takeaways for Industry”, Skadden Publication, 30 July 2025.

[6] “Winning the Race AMERICA’S AI ACTION PLAN”, Executive Office of the President, United States of America, July 2025, p. 15.

[7] Ibid., p. 14.

[8] “White House Unveils America’s AI Action Plan”, The White House, 23 July 2025.

[9] “The Trump Administration’s 2025 AI Action Plan – Winning the Race: America’s AI Action Plan – and Related Executive Orders”, Sidley, 30 July 2025.

[10] Vivek Chilukuri et al., “Noteworthy | America’s AI Action Plan”, CNAS, 23 July 2025.

[11] “AI Action Plan-Pillar 3”, AI.Gov, July 2025; Pablo E. Carrillo et al., “Key Insights on President Trump’s New AI Executive Order and Policy & Regulatory Implications”, Squire Patton Boggs, February 2025.

[12] Executive Orders, “Promoting the Export of the American AI Technology Stack”, The White House, 23 July 2025.

[13] Sorelle Friedler et al., “What to Make of the Trump Administration’s AI Action Plan”, Brookings, 23 July 2025.

[14] “Preventing Woke AI in the Federal Government”, Executive Order, The White House, 23 July 2025.

[15] Matt O’Brien and Sarah Parvini, “Trump Signs Executive Order on Developing Artificial Intelligence ‘Free from Ideological Bias’”, AP News, 24 January 2025; “Winning the AI Race: America’s AI Action Plan-Pillar 1”, The White House, July 2025, p. 4.

[16] Amos Toh, “How Trump’s AI Policy Could Compromise the Technology”, Brennan Center for Justice, 1 August 2025.

[17] “Preventing Woke AI in the Federal Government”, Executive Order, The White House, 23 July 2025.

[18] “NSF Releases List of Terminated Grants”, COSSA Washington Update, 27 May 2025.

[19] “Inside Trump’s Ambitious AI Action Plan”, HAI, 24 July 2025.

[20] “Federal Research and Development (R&D)  Funding: FY2026”, Congressional Research Service, 3 September 2025.

[21] Lisa Larrimore Ouellette, “The Trump Administration’s Multi-Front Assault on Federal Research Funding”, Just Security, 9 July 2025.

[22] “Promoting the Export of the American AI Technology Stack”, The White House, 23 July 2025.

[23] “Memorandum of Cooperation Regarding the Technology Prosperity Deal Between the Government of the United States of America and the Government of Japan”, The White House, 28 October 2025.

[24] “What the US AI Action Plan Means for Export Controls and US National Security”, Alvarez and Marsal, 23 September 2025.