By Micah Drayton and Crystel Saraie.
In May 2025, the U.S. Copyright Office (USCO) released a pivotal report addressing the application of the “fair use” doctrine to the training of generative AI models. This report, the third in a series examining the intersection of copyright law and artificial intelligence, scrutinises whether using copyrighted materials to train AI systems qualifies as fair use under current legal standards.
What Is Fair Use and How Is It Determined?
Fair use is a legal doctrine that allows limited use of copyrighted material without the rights holder’s permission, primarily for purposes such as commentary, criticism, news reporting, teaching, scholarship, or research. Whether a particular use qualifies as “fair” is determined on a case-by-case basis using a four-factor test set out in Section 107 of the U.S. Copyright Act:
- The Purpose and Character of the Use – Courts consider whether the use is commercial or non-commercial and whether it is “transformative” (i.e., does it add new meaning or purpose to the original work?).
- The Nature of the Copyrighted Work – This factor assesses whether the original work is more factual (which leans toward fair use) or highly creative (which offers stronger protection to the creator).
- The Amount and Substantiality of the Portion Used – This looks at both the quantity and qualitative value of the material used. Using the “heart” of a work, even if small, may weigh against fair use.
- The Effect of the Use on the Potential Market for or Value of the Work – Courts evaluate whether the new use could harm the market for the original or serve as a substitute.
These factors are weighed together, not applied in isolation, and the outcome depends on the specific circumstances of each case.
Key Findings of the USCO Report
The USCO’s analysis suggests that while AI training might be considered transformative in certain contexts, such as research or non-commercial applications, the large-scale, commercial use of copyrighted works for training generative AI models likely exceeds the bounds of fair use. [1] The report emphasises that such practices could directly compete with the original works, potentially undermining the market for the creators’ content.
The report also challenges the analogy often made between AI learning and human learning, asserting that the scale and nature of AI data ingestion differ significantly from human cognitive processes. This distinction weakens the argument that AI training should be treated similarly to human learning under the fair use doctrine.
Industry and Legal Repercussions
The USCO’s stance has significant implications for the tech industry, particularly companies developing generative AI technologies. By questioning the applicability of fair use in AI training, the report opens the door for increased scrutiny and potential legal challenges against AI developers who utilise copyrighted materials without authorisation.
The report’s release was followed by notable administrative changes, including the dismissal of Shira Perlmutter, the Register of Copyrights, and Carla Hayden, the Librarian of Congress. [3] [2] These actions have sparked concerns about political interference in the USCO’s operations and the broader implications for copyright enforcement in the digital age.
Looking Ahead
As the legal landscape continues to evolve, the USCO’s report serves as a critical reference point for ongoing debates about the balance between fostering innovation in AI and protecting the rights of content creators. Stakeholders across the creative and tech industries will need to navigate these complex issues as they develop and implement AI technologies in compliance with copyright laws.
This publication is distributed with the understanding that the author, publisher, and distributor of this publication and/or any linked publication are not rendering legal, accounting, or other professional advice or opinions on specific facts or matters and, accordingly, assume no liability whatsoever in connection with its use. Pursuant to applicable rules of professional conduct, portions of this publication may constitute Attorney Advertising. The choice of a lawyer is an important decision and should not be based solely upon advertisements.
[1] Edmonds, L. (n.d.). The US Copyright Office has thoughts on how AI is trained. Big Tech may not like it. Business Insider. https://www.businessinsider.com/ai-training-copyright-laws-big-tech-fair-use-openai-meta-2025-5
[2] Fried, I. (2025, May 13). U.S. Copyright Office’s AI report Sparks new fight. AI copyright report sparks new fight. https://www.axios.com/2025/05/13/ai-copyright-office-report-fight
[3] O’Brien, M. (2025, May 12). Trump administration fires top copyright official days after firing librarian of Congress. AP News. https://apnews.com/article/copyright-director-firing-government-trump-7ab99992a96131bce7de853b66feec68