The Shortcomings of United States Artificial Intelligence Regulations: An International Comparison
With the increasing prevalence of artificial intelligence (AI), there is a simultaneous need for comprehensive regulations. Almost every app, website, search engine, and social media has adopted its own generative AI system. Although these systems have the power to reduce operational costs, increase efficiency and productivity, and enhance innovation, generative AI introduces a whirlwind of risks as well. From data leaks and privacy concerns to harmful deepfakes and the circulation of inaccurate information, AI regulation is needed to limit the potential harms resulting from its use.[1] Unfortunately, shifting political priorities has proven to be a significant obstacle for America, complicating efforts to establish a unified AI policy framework like other regions.
In March 2024, the United Nations (UN) General Assembly adopted its first-ever AI resolution.[2] Despite being non-binding, it demonstrates global acknowledgement of the pervasive need for AI regulations.[3] All 193 UN member states endorsed the resolution, which was led by the United States and co-sponsored by 123 other member states.[4] Moreover, the UN resolution provides a guideline to frame other regulations after, both in a national and international context.[5] In recent years, the spectrum of AI regulatory measures across countries have ranged from non-existent to detailed and prescriptive, like the European Union (EU) AI Act.[6]
The EU AI Act implemented a tiered approach to regulating general purpose AI models.[7] Larger companies, whose systems present higher risks, are subjected to more stringent obligations, whereas smaller companies with lower-risk systems are afforded greater flexibility.[8] By tailoring the regulatory requirements to correspond with the risk level of the system, the EU AI Act aims to foster innovation among emerging companies with lower-risk profiles.[9] High-risk AI systems include those used in education, healthcare, law enforcement, border management, and elections.[10] This comprehensive regulatory framework for combatting the risks of AI places the EU at the forefront of global attempts to regulate AI.[11]
The pressure is on for the United States to implement an equally, if not more, thorough oversight system. President Biden marked a “critical step in defining AI regulation and accountability across multiple sectors”[12] when he issued White House Executive Order (EO) 14110 on AI in 2023.[13] Despite discussing the goals and risks of AI, EO 14110 did not prescribe clear mandates to mitigate AI risks.[14] With that said, President Trump has now revoked most of the AI regulation introduced by President Biden.[15]
President Trump has emphasized his goal of “maintaining U.S. leadership in AI.”[16] Unlike the Biden administration, which placed a significant focus on risk mitigation in the context of AI, the Trump administration has adopted a “market-oriented approach that emphasizes fostering innovation and countering the growing technological influence of China.”[17] Over the next four years, the Trump administration is expected to utilize AI “as a crucial technology for enhancing national security and economic competitiveness.”[18]
The United States’ approach to AI regulation remains fragmented, hindered by shifting political priorities and the absence of a cohesive federal framework. While global initiatives like the UN resolution and the EU AI Act provide structured oversight, America oscillates between prioritizing risk mitigation and innovation with each administration. Furthermore, the current patchwork state laws create compliance challenges for businesses and risk undermining long-term AI governance.[19] To maintain leadership while addressing critical risks, the United States must establish a balanced, bipartisan framework that fosters innovation, ensures transparency, and remains adaptable beyond political cycles.
Juliette Donovan is a staff member of Fordham International Law Journal Volume XLVIII.
[1] Sabrina Ortiz, The UN’s AI Resolution is Non-binding, but Still a Big Deal – Here’s Why, ZDNet (Mar. 22, 2024, 1:25 PM), https://www.zdnet.com/article/uns-ai-resolution-is-non-binding-but-still-a-big-deal-heres-why/.
[2] G.A Res. A/78/L.49 (Mar. 21, 2024).
[3] Id.; Ortiz, supra note 1.
[4] Ortiz, supra note 1.
[5] Id.
[6] Ron Babin, Global AI Regulations: Beyond the U.S. and Europe, CIO (Nov. 21, 2024), https://www.cio.com/article/3608168/global-ai-regulations-beyond-the-u-s-and-europe.html.
[7] Cathy Li, UN and EU Both Agree on New AI Rules, World Economic Forum (Sep. 10, 2024), https://www.weforum.org/stories/2024/04/artificial-intelligence-technology-news-april-2024/.
[8] Id.
[9] Id.
[10] Shiona McCallum et al., MEPs Approve World’s First Comprehensive AI Law, BBC (Mar. 13, 2024), https://www.bbc.com/news/technology-68546450.
[11] Id.
[12] Aaron Klein et al., 1 Year Later, How Has the White House AI Executive Order Delivered on its Promises?, Brookings (Nov. 4, 2024), https://www.brookings.edu/articles/one-year-later-how-has-the-white-house-ai-executive-order-delivered-on-its-promises/.
[13] Exec. Order No. 14110, 88 Fed. Reg. 75191 (Nov. 1, 2023).
[14] Klein, supra note 12.
[15] See AI Legislation in the US: A 2025 Overview, Software Improvement Group (Jan. 24, 2025), https://www.softwareimprovementgroup.com/us-ai-legislation-overview/.
[16] Bae Kim & Lee LLC, AI Regulation Direction of President Trump’s Second-Term Administration, Lexology (Jan. 31, 2025), https://www.lexology.com/library/detail.aspx?g=c7c12b03-93c1-45dd-83c3-33d3b1642540.
[17] Id.
[18] Id.
[19] See Keely Quinlan, There’s a Wave of State AI Legislation Coming, New Report Says, Statescoop (Oct. 24, 2024), https://statescoop.com/ai-legislation-state-regulation-2024/#:~:text=In%20lieu%20of%20a%20federal,legislation%20to%20come%20in%202025.
This is a student blog post and in no way represents the views of the Fordham International Law Journal.