On January 23, 2025, US President Donald Trump signed an executive order titled "Removing Barriers to American Leadership in Artificial Intelligence," marking a significant shift in U.S. AI policy. This order revokes and replaces key elements of the Biden administration's approach to AI governance.
Here's a simple and comprehensive breakdown of the executive order's main provisions in this quick insight.
Core Framework and Purpose
President Trump signed the executive order "Removing Barriers to American Leadership in Artificial Intelligence" with the fundamental aim of maintaining U.S. leadership in AI innovation through free markets, research institutions, and entrepreneurial spirit. The order explicitly establishes a policy directive to enhance America's global AI dominance for promoting human flourishing, economic competitiveness, and national security.
Definition of Artificial Intelligence Relied Upon
The order refers to a specific definition of artificial intelligence based on 15 USC § 9401(3):
The term “artificial intelligence” means a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments. Artificial intelligence systems use machine and human-based inputs to— (A) perceive real and virtual environments; (B) abstract such perceptions into models through analysis in an automated manner; and (C) use model inference to formulate options for information or action.
The definition specifies that AI systems must use both machine and human-based inputs to perform three key functions:
Perceive real and virtual environments
Abstract these perceptions into models through automated analysis
Use model inference to formulate options for information or action
Significance and Context of the AI Definition
This definition choice is notable for some reasons:
The order maintains continuity with existing legal frameworks by using the established definition from the National Artificial Intelligence Initiative Act of 2020.
However, this definition has limitations - it may not clearly encompass generative AI unless broadly interpreted, yet such broad interpretation could include basic computer systems not typically considered AI.
The definition's scope is particularly relevant as it serves as the foundation for all provisions and implementations under the new executive order, including the development of the mandated AI action plan within 180 days.
Implementation Structure
The order creates an implementation framework centered around key officials including David Sacks as Special Advisor for AI and Crypto, working alongside the Assistant to the President for Science and Technology and National Security Affairs.
Within 180 days, these officials must develop a comprehensive AI action plan, coordinating with economic policy advisors, domestic policy advisors, the OMB Director, and relevant agency heads.
Regulatory Changes and Review Process
A significant aspect of the order is its systematic dismantling of previous AI governance structures. The Office of Management and Budget has been given 60 days to revise Memoranda M-24-10 and M-24-18, which were cornerstone AI policy documents under the Biden administration. Agency heads are directed to suspend, revise or rescind actions that conflict with the new policy direction. Nik Marda, who had formerly worked in the Office of Science and Technology Policy at the Biden Administration, had remarked:
"That's not only bad policy given the many real bias and discrimination risks from AI, but it's also a significant departure from both the bipartisan law that mandates this guidance and Trump's own 2020 executive order on AI — both of which name civil rights as an objective for federal use of AI."
Now, for some context: the OMB Memoranda M-24-10 and M-24-18 were two critical policy documents issued in 2024 under the Biden administration that established the federal government's AI governance framework.
M-24-10: Core Governance Framework
Released in March 2024, Memorandum M-24-10 "Advancing Governance, Innovation, and Risk Management for Agency Use of Artificial Intelligence" had established fundamental AI governance structures across federal agencies. Key elements included:
Creation of Chief AI Officer (CAIO) positions in federal agencies
Establishment of the Chief AI Officer Council (CAIOC) for interagency coordination
Requirements for agencies to develop compliance plans for AI governance
Framework for identifying and managing "rights-impacting" and "safety-impacting" AI systems
To note, M-24-10 never included common commercial products with relevant embedded AI functionality.
M-24-18: AI Acquisition Guidelines
Issued on October 3, 2024, Memorandum M-24-18 "Advancing the Responsible Acquisition of Artificial Intelligence in Government" focused specifically on federal AI procurement. The memorandum:
Set requirements for contracts awarded after March 23, 2025
Required agencies to disclose if AI systems were rights-impacting or safety-impacting
Mandated 72-hour reporting of serious AI incidents
Established contractor requirements for AI system documentation and training data transparency
Implementation Status of the Memoranda
Before Trump's executive order, agencies were actively implementing these memoranda. For example:
The USDA had developed and published its compliance plan in September 2024
Federal agencies were required to identify contracts involving rights-impacting AI by November 1, 2024
The framework was being integrated with national security considerations through the October 2024 National Security Memorandum on AI
Trump's January 23, 2025 executive order now requires OMB to revise both memoranda within 60 days. Most probably, it could be a significant shift away from Biden's emphasis on AI safety and governance toward reduced regulatory oversight.
Ideological Shift in AI Development
The order marks a stark departure from previous approaches by emphasizing AI systems "free from ideological bias or engineered social agendas". This represents a fundamental shift from the Biden administration's focus on AI safety and ethical considerations. However, the ideological shift may also signal that this administration is not interested in DEI-related and over-regulatory Responsible AI ideas, which have been under consideration in the previous administration. It must be noted that the previous administration's obsession with a few companies to merely focus on AI innovation & AI safety was also counterproductive, despite the fact that the main Executive Order of the Biden administration keenly focused on capacity building on relevant responsible AI measures for their own government personnel.
Energy Infrastructure Expansion
Trump declared a national energy emergency to fast-track the approval of energy projects critical for AI development. During his address at the World Economic Forum in Davos, Switzerland, he emphasized that the U.S. must double its current energy capacity to meet the demands of AI technologies. This includes constructing power plants specifically dedicated to supporting AI data centres.
These facilities will not be subject to traditional climate objectives, with Trump allowing them to use any fuel source, including coal as a backup. He highlighted the reliability of coal, describing it as "a great backup" due to its resilience against weather and other disruptions.
The president also proposed a co-location model, where power plants are built directly adjacent to AI data centers. This approach bypasses reliance on the traditional electricity grid, which Trump described as "old and vulnerable." By connecting power plants directly to data centers, companies can ensure uninterrupted energy supply for their operations.
The Stargate Initiative
In conjunction with these measures, Trump announced the Stargate Initiative, a joint venture between OpenAI, Oracle, and SoftBank, aimed at investing up to $500 billion over four years in building AI infrastructure. The initiative begins with a $100 billion investment in data centers, starting with a flagship facility in Texas. The project is supposed to be expected to create over 100,000 jobs and significantly enhance U.S. computing capabilities for AI.
Industry Impact
The executive order also introduces sweeping changes aimed at reducing regulatory oversight on AI companies, which Trump has characterised as "unnecessarily burdensome" under Biden's administration.
Deregulation
Trump's order repealed Biden's 2023 executive order that mandated stricter oversight of AI development. Key provisions eliminated include:
Requirements for AI developers to submit safety testing data before deploying high-risk technologies.
Obligations for federal agencies to assess and mitigate potential harms caused by government use of AI tools.
Mandatory transparency measures for companies building powerful AI models.
By removing these guardrails, the Trump administration seeks to foster innovation by reducing compliance costs and accelerating deployment timelines. This deregulatory approach aligns with Trump's broader economic agenda of empowering private sector-led growth in technology industries.
What will happen to the AI Diffusion Export Control Rule?
Reevaluation of Licensing Requirements
The AI Diffusion Export Control Rule currently imposes global licensing requirements for exporting advanced computing integrated circuits (ICs) and AI model weights trained with more than 10 to the power of 26 computational operations. These restrictions are designed to prevent U.S. adversaries, particularly China, from accessing cutting-edge AI technologies while allowing controlled exports to trusted allies under specific license exceptions.Trump's executive order, which emphasises reducing regulatory barriers and fostering U.S. innovation, may lead to:
Relaxation of Licensing Rules: The administration could ease licensing requirements for U.S. companies exporting AI technologies to middle-tier countries (e.g., India, Brazil) to boost international market competitiveness.
Focus on Deregulation: There is a strong likelihood that Trump will reduce compliance burdens for U.S. firms by eliminating or simplifying reporting and monitoring obligations tied to export licenses.
Strategic Adjustments to Country Tiers
The Biden administration's framework divides countries into three tiers:
Top-tier allies (e.g., NATO members, Japan, South Korea) enjoy unrestricted access.
Middle-tier nations (e.g., India, Saudi Arabia) face quotas on computing power imports.
Adversarial nations (e.g., China, Russia) remain blocked from accessing advanced U.S. AI technologies.
Trump's transactional foreign policy approach suggests potential adjustments:
Reclassification of Countries: Middle-tier nations like India and Israel could be moved into the top tier as part of bilateral negotiations or geopolitical alliances.
Use as Diplomatic Leverage: Trump may leverage access to U.S. AI technologies as a bargaining chip in trade agreements or security partnerships.
Private Sector Benefits
The rollback of regulations has been welcomed by industry leaders who argue that excessive oversight stifles innovation and hinders competitiveness. For example, OpenAI CEO Sam Altman has praised the administration's focus on enabling rapid development of computing infrastructure through initiatives like Stargate.
Trump’s administration has framed these changes as essential for maintaining U.S. leadership in AI while countering international competitors like China. The emphasis on deregulation reflects a shift toward prioritizing economic competitiveness over ethical or safety considerations.
Conclusion
The January 23 executive order demonstrates President Trump's aggressive push for U.S. dominance in artificial intelligence by combining massive infrastructure investments with significant deregulation of the industry. While these measures promise economic growth and technological advancement, they come with potential risks related to environmental impact and reduced oversight.
If you have any questions or need assistance regarding this policy or its implications, feel free to contact us at vligta@indicpacific.com.
Thanks for reading this insight.
Since May 2024, we have launched some specialised practice-based technology law and artificial intelligence & law training programmes at indicpacific.com/train.
We offer special discounts for technology teams who may be interested in enrolling for the training programmes on bulk. Feel free to choose your training programme at indicpacific.com/train and contact us at vligta@indicpacific.com.
Comments