The U.S. Copyright Office’s report, published on January 29, 2025, examines the copyrightability of outputs created using generative artificial intelligence (AI). The report is part of a broader initiative launched in 2023 to analyze copyright law and policy issues related to AI, following public listening sessions, webinars, and over 10,000 comments received in response to a notice of inquiry. Below are the key points, with a focus on aspects relevant to AI-generated music and human interaction:

1. Core Principle of Human Authorship:
   • Copyright law requires human authorship for a work to be eligible for protection. Purely AI-generated works, where a machine determines the expressive elements without sufficient human control, are not copyrightable. This is rooted in the constitutional goal of copyright to incentivize human creativity, not machine outputs.
2. AI as a Tool vs. Stand-In for Creativity:
   • When AI is used as a tool to assist human creativity (e.g., enhancing, editing, or generating drafts that humans refine), the resulting work can be copyrightable, provided the human contribution is significant and creative. However, if AI acts as a “stand-in” for human creativity, generating outputs without meaningful human input, the work is not protectable.
3. Prompts and Human Control:
   • Merely providing prompts to an AI system, even detailed ones, does not currently constitute sufficient human control to claim authorship. The report highlights the unpredictability of AI outputs, noting that prompts do not reliably determine the expressive elements of the final work. For example, a prompt requesting a “jazz song in the style of Miles Davis” might produce a song, but the user’s lack of control over specific musical elements (melody, harmony, rhythm) means the output is not copyrightable.
4. Expressive Inputs and Modifications:
   • Copyright protection may apply if a human provides “expressive inputs” (e.g., a hand-drawn sketch or a pre-existing musical composition) that are perceptible in the AI-generated output. Similarly, if a human significantly modifies or arranges AI-generated material, the modified work can be copyrightable, provided the modifications meet the minimum standard of originality. For instance, a human arranging AI-generated musical snippets into a cohesive song could claim copyright over the arrangement, but not the raw AI outputs.
5. Case-by-Case Analysis:
   • The copyrightability of AI-generated works is determined on a case-by-case basis, focusing on the degree of human control and creativity. The report emphasizes that existing copyright law is flexible enough to handle AI-related issues without requiring legislative changes, though future technological advancements might necessitate revisiting this stance.
6. Specific to Music:
   • For AI-generated music, the report implies that copyright protection is possible if a human author contributes perceptible creative elements. Examples include composing lyrics to accompany an AI-generated melody, arranging AI-generated musical fragments into a new structure, or using AI to enhance a human-authored composition (e.g., adding effects or refining instrumentation). In such cases, the human-authored elements are protectable, but purely AI-generated portions must be disclaimed in copyright applications.
7. Examples and Precedents:
   • The report references cases like Thaler v. Perlmutter (2023), where a court upheld the Copyright Office’s refusal to register a purely AI-generated image, reinforcing that human authorship is essential. It also cites examples like Zarya of the Dawn, where a comic book’s human-authored text and arrangement of AI-generated images were copyrightable, but the AI images themselves were not.
8. Policy Recommendations:
   • The Copyright Office concludes that no new laws are needed at this time, as existing principles can address AI-generated works. However, it acknowledges that future AI advancements (e.g., systems allowing greater user control over outputs) might change this analysis. The Office also plans to update its registration guidance and practices to clarify disclosure requirements for AI-generated content.
9. Broader Implications:
   • The report underscores the importance of transparency in AI use, encouraging creators to disclose AI contributions to ensure accurate copyright registration. It also highlights potential economic and cultural impacts, such as the risk of AI-generated works competing with human-authored works in creative industries like music, film, and literature.