Introduction
The Trump administration has introduced two executive orders that significantly reshape the federal government’s approach to artificial intelligence (AI) regulation [4]. These orders prioritize innovation and competition, moving away from the previous administration’s focus on safety and individual rights [4]. The new directives aim to reduce bureaucratic barriers, foster AI development [4], and enhance US leadership in the sector [4].
Description
President Trump has issued two executive orders that significantly alter the federal government’s approach to artificial intelligence (AI) regulation [4], emphasizing innovation and competition while moving away from the previous administration’s focus on safety and individual rights. The new directives prioritize reducing bureaucratic barriers to foster AI development and enhance US leadership in the sector [4], allowing federal agencies to become more agile and efficient in their adoption of emerging technologies.
These executive orders revoke the Biden administration’s Executive Order 14110 [4], which aimed to ensure safe and trustworthy AI [4], and mandate a comprehensive review of Biden-era AI policies [4]. Senior officials are tasked with identifying actions that conflict with the new strategy and recommending modifications or rescissions [4]. The Office of Management and Budget (OMB) has released two significant memos [5] [6] [7], M-25-21 and M-25-22 [5] [7], which outline a new federal approach to the use and procurement of AI systems [5], reflecting a shift towards innovation and competitiveness while emphasizing data privacy [5], public trust [5], and operational excellence [5].
M-25-21 establishes minimum risk management practices for “high-impact AI,” which includes pre-deployment testing [7], AI impact assessments [7], and ongoing monitoring [7]. This memo defines “high-impact AI” as systems that significantly affect civil rights [7], privacy [2] [3] [5] [7], human health [7], and critical infrastructure [7], ensuring that AI use cases with significant implications are subject to enhanced scrutiny [2]. Agencies are now required to maintain inventories of AI use cases and document compliance within one year, discontinuing any noncompliant high-impact applications [3]. Additionally, agencies must appoint Chief AI Officers (CAIOs) to lead AI initiatives, evaluate AI maturity [5], identify growth opportunities [5], and manage risks associated with high-impact AI systems [2] [5]. They are also required to report annually on AI use cases [5], particularly those with significant legal or safety implications [5].
Meanwhile, M-25-22 focuses on federal procurement of AI [7], emphasizing performance-based contracting [5] [7], ongoing testing [7], and monitoring of AI systems [7], while ensuring the protection of contractors’ intellectual property rights and proper use of government data in AI applications [7]. The updated policies streamline the procurement process and reduce reporting burdens [2], promoting a competitive marketplace while safeguarding privacy and lawful data use [2]. A notable requirement is the formation of cross-functional acquisition teams that include experts from various fields to assess the potential impact of AI systems [5], particularly in high-risk areas [5]. The General Services Administration (GSA) will create a web-based repository to facilitate AI procurement and share best practices [6].
The Trump administration’s approach includes establishing AI governance measures [4], such as appointing CAIOs and developing enterprise AI strategies [4], while also expanding acquisition policies to ensure interoperability and prevent vendor lock-in [4]. An AI Action Plan is to be developed within 180 days [4], focusing on maintaining AI leadership and minimizing regulatory hurdles [4]. The new policy allows for greater discretion in managing risks and mandates public feedback mechanisms for high-impact AI use cases [3]. However, concerns have been raised regarding the impact of workforce reductions on agencies’ ability to implement these requirements effectively [3].
While the new orders represent a rollback of Biden-era policies [4], they do not indicate complete deregulation [4]. Instead, they emphasize innovation and global dominance over risk management [4]. The focus on reducing compliance burdens suggests that AI developers may face fewer federal reporting and documentation requirements [4], particularly regarding risks associated with AI systems [4]. The Department of Government Efficiency’s use of AI has drawn scrutiny for potentially bypassing privacy and security protocols [3], particularly in sensitive areas like federal employment decisions and fraud detection [3].
Despite the federal shift [1] [4], state-level AI regulation is expected to continue or expand [4], potentially leading to a fragmented regulatory landscape [4]. The Trump administration has maintained strict export controls on AI-related technologies [4], particularly concerning semiconductor technology [4], amid concerns about foreign competition [4].
The EU AI Act imposes compliance obligations on AI providers and deployers [4], contrasting with the Trump administration’s stance against foreign regulations perceived as barriers to innovation [4]. This divergence may complicate compliance for US AI companies operating in both jurisdictions [4], as they navigate conflicting expectations regarding content moderation and misinformation controls [4]. The Trump administration’s emphasis on free speech in AI governance presents challenges [4], particularly in light of the EU’s regulatory approach to misinformation [4]. This growing divide underscores the need for US companies to carefully evaluate their AI deployment strategies in light of differing global standards [4].
Additionally, significant government-backed initiatives [4], such as the Stargate Initiative [4], aim to enhance US AI capabilities [4], while efforts are underway to develop AI tools for federal government use [4]. The effectiveness of the new guidance will depend on how well the Office of Management and Budget collaborates with federal agencies to ensure compliance with transparency and risk management standards [3]. Federal agencies are encouraged to adopt effective AI solutions to benefit the public [6], signaling increased investment opportunities for managed service providers (MSPs) that support federal clients [6]. MSPs can assist clients by identifying high-impact AI use cases [6], conducting risk assessments [5] [6], supporting the appointment of CAIOs [6], and developing AI strategies for those in government supply chains [6]. They can also review or draft AI-related contract clauses [6], recommend open formats to avoid vendor lock-in [6], and help track AI performance post-deployment [6]. Additionally, MSPs can assess data governance readiness [6], advise on AI infrastructure [6], monitor AI model effectiveness [6], and provide training for staff on AI literacy and governance [6].
Conclusion
The Trump administration’s executive orders mark a significant shift in AI regulation, prioritizing innovation and competition over stringent risk management. While these changes aim to enhance US leadership in AI, they also present challenges, particularly in terms of compliance with international standards and the potential for a fragmented regulatory landscape. The success of these initiatives will depend on effective collaboration between federal agencies and the private sector, as well as the ability to balance innovation with necessary safeguards.
References
[1] https://timesofindia.indiatimes.com/technology/tech-news/donald-trumps-new-ai-directive-removes-two-biden-era-restrictions-agencies-must-adopt-a-forward-leaning-/articleshow/120095427.cms
[2] https://www.presidency.ucsb.edu/documents/white-house-fact-sheet-eliminating-barriers-for-federal-artificial-intelligence-use-and
[3] https://federalnewsnetwork.com/artificial-intelligence/2025/04/trumps-ai-policy-shifts-focus-to-high-impact-use-cases/
[4] https://www.jdsupra.com/legalnews/trump-administration-early-executive-7758216/
[5] https://www.biometricupdate.com/202504/trump-administration-announces-overhaul-of-federal-ai-policy
[6] https://compliancescorecard.com/2025/04/federal-ai-policy-update-executive-orders-reshape-government-ai-approach/
[7] https://www.wiley.law/alert-Trump-Administration-Revamps-Guidance-on-Federal-Use-and-Procurement-of-AI
