Understanding the Copyright Office’s AI Initiative
The U.S. Copyright Office launched a comprehensive initiative in early 2023 to examine the copyright law and policy issues raised by artificial intelligence (AI), including the scope of copyright in AI-generated works and the use of copyrighted materials in AI training. After hosting public listening sessions and webinars, the Office published a notice of inquiry in the Federal Register in August 2023, which received over 10,000 comments by December 2023.
The Office is issuing its analysis in a multi-part Report, with each part published as completed. Part 1 was published on July 31, 2024, addressing digital replicas. Part 2 was published on January 29, 2025, addressing the copyrightability of outputs created using generative AI. The forthcoming Part 3 will address the legal implications of training AI models on copyrighted works, licensing considerations, and the allocation of any potential liability.
This initiative represents the most comprehensive official examination of AI and copyright issues undertaken by the federal agency responsible for administering U.S. copyright law. The extensive public input process, including thousands of comments from stakeholders across creative industries, technology companies, and legal professionals, has informed the office’s analysis and recommendations.
Part 1: Digital Replicas and Voice Likeness Protection
The first part of the Copyright Office’s AI report, released in July 2024, focused specifically on digital replicas—AI-generated content that realistically depicts individuals without their consent. This portion of the report recommended federal legislation to address the unauthorized distribution of digital replicas that falsely depict individuals, responding to growing concerns about deepfakes and AI-generated impersonations.
The digital replica issue represents one of the most immediate policy challenges raised by AI technology, as current copyright law provides limited protection against unauthorized AI-generated likenesses and voice replicas. The Copyright Office’s recommendation for federal legislation acknowledges that existing legal frameworks may be insufficient to address these emerging threats to individual privacy and rights of publicity.
This first report established the foundation for the office’s broader examination of AI copyright issues while addressing one of the most pressing concerns raised by stakeholders during the public comment process. The focus on digital replicas also reflects the office’s recognition that different aspects of AI technology may require distinct legal approaches and policy solutions.
Part 2: Copyrightability of AI-Generated Content
The second part of the report, released on January 29, 2025, addresses the copyrightability of outputs created using generative AI. The Copyright Office affirms that existing principles of copyright law are flexible enough to apply to this new technology, as they have applied to technological innovations in the past. It concludes that the outputs of generative AI can be protected by copyright only where a human author has determined sufficient expressive elements.
The Human Authorship Requirement
The Copyright Office maintains that human authorship and creativity remain essential in the quest to obtain copyright protection for works involving materials created by artificial intelligence. The Office further notes that no U.S. court has recognized copyrights in materials created by non-humans, and courts that have spoken on this issue have directly rejected the possibility.
For any work to be copyrightable in the U.S., the Office affirms that the human authorship must contain some degree of originality and cannot be merely the result of time and effort. With this in mind, along with noting that each determination of protectability must be made on a case-by-case basis, the Office lays out three scenarios in which AI-generated works may qualify for copyright protection.
The office’s position on human authorship represents continuity with longstanding copyright principles while addressing new technological challenges. This requirement reflects both constitutional foundations of copyright law and practical considerations about the purposes and incentives that copyright protection is intended to serve.
Three Scenarios for AI Copyright Protection
According to the Copyright Office, the use of AI tools to assist rather than stand in for human creativity should not affect the availability of copyright protection for the output. Among the examples referenced by the Office include the longstanding use of computer-assisted tools in the creative process (such as “color correction, detail sharpening, or de-blurring”) as well as the use of generative AI tools for “song ideation” or a “preliminary outline for literary works,” where the AI output is not directly incorporated into the human-authored work.
The office identifies three specific scenarios where AI-involved works may still qualify for copyright protection. First, assistive uses of AI tools that enhance human creativity rather than replace it should not affect copyrightability. Second, situations where human-authored work remains perceptible in the AI output may support copyright claims. Third, cases where humans make creative arrangements or modifications of AI output may establish sufficient human authorship for protection.
The Office concluded that under existing law, “purely AI-generated material” created by entering prompts is not copyrightable. Its analysis relied on Thaler v. Perlmutter, which upheld the Office’s refusal to register a claimed copyright in an image “autonomously created by a computer algorithm running on a machine.”
Registration Guidance and Current Practice
The Copyright Office has issued registration guidance for works containing AI-generated content, providing practical direction for creators seeking to register works that involve AI assistance. This guidance helps applicants understand how to describe AI involvement in their registration applications and what level of human creativity is necessary to support copyright claims.
Consistent with the principle that only human-created works are eligible for copyright protection, the Office has refused to register multiple works generated by AI platforms. Notable cases include the office’s rejection of applications for works like “Théâtre D’opéra Spatial,” which was created using the Midjourney AI platform with extensive prompting but determined to contain more than a de minimis amount of AI-generated content.
The practical application of these principles continues to evolve as the office examines individual registration applications and develops clearer standards for evaluating human authorship in AI-assisted works. The case-by-case approach reflects the complexity of determining human creative contribution in various AI collaboration scenarios.
Stakeholder Engagement and Public Input
The Copyright Office has conducted extensive stakeholder engagement, hosting public listening sessions and webinars, meeting with experts and stakeholders, and reviewing more than 10,000 responsive comments that served to inform the conclusions in the reports. These listening sessions covered various creative sectors, including visual arts, literary works including software, audiovisual works, and music and sound recordings.
The public comment process revealed significant interest and diverse perspectives on AI copyright issues. Approximately half of the more than 10,000 comments that the Office received in response to the Notice of Inquiry addressed copyrightability questions. This extensive input has helped shape the office’s understanding of practical challenges facing different creative industries and technology sectors.
The ongoing engagement reflects the office’s recognition that AI technology continues evolving rapidly and that effective policy development requires sustained dialogue with affected communities. The office has committed to continued monitoring of technological developments and their implications for copyright law and policy.
Looking Ahead: Part 3 and Future Developments
The final, forthcoming Part 3 will address the legal implications of training AI models on copyrighted works, including licensing considerations and the allocation of any potential liability. This upcoming report is highly anticipated as it will address one of the most contentious aspects of AI development—the use of copyrighted materials in training datasets.
The training data issue involves complex questions about fair use, licensing requirements, and liability for AI developers and users. Many stakeholders have raised concerns about unauthorized use of copyrighted works in AI training, while others argue that such use may qualify for fair use protection under certain circumstances.
The vast majority of those who submitted comments during the public comments period urged the office to protect copyrighted works from being used by corporations to train generative AI without consent and without crediting and compensating creators. This input suggests Part 3 will need to address significant stakeholder concerns about unauthorized use of creative works in AI development.
Real-World Applications and Current Disputes
The practical implications of the Copyright Office’s positions are already playing out in ongoing legal disputes and registration decisions. In September 2024, Jason Allen filed a declaratory judgment action under the Administrative Procedure Act against the Copyright Office in the District of Colorado, seeking to overturn the Office’s refusal to register “Théâtre D’opéra Spatial.” Allen had used extensive prompting, Adobe Photoshop corrections, and AI upscaling tools in creating the work.
These cases help establish practical precedents for how copyright law applies to AI-generated content and clarify the boundaries between human and machine creativity in copyright analysis. The outcomes of such disputes will likely influence how creators, technology companies, and legal professionals approach AI copyright issues in the future.
The office’s case-by-case approach to registration decisions provides ongoing guidance about what level of human involvement is sufficient to support copyright claims when AI tools are involved in the creative process.
Congressional Oversight and Policy Development
The Committee on House Administration has broad oversight authority over the Library of Congress, which includes the U.S. Copyright Office. In 2024, Committee Chairman Bryan Steil sent a letter to the Copyright Office requesting updates on the AI reports, noting that “creators, technologists, and stakeholders from across the spectrum are relying on clear, authoritative guidance from your office to navigate this evolving landscape.”
Congressional interest in the Copyright Office’s AI initiative reflects the broader policy significance of these issues and the need for clear federal guidance on AI copyright questions. The oversight process helps ensure that the office’s work addresses the practical needs of affected industries and provides timely guidance for ongoing policy development.
The intersection of AI technology and copyright law involves complex technical, legal, and policy considerations that require careful analysis and stakeholder input. The Copyright Office’s systematic approach to examining these issues through its multi-part report process demonstrates the seriousness with which federal policymakers are approaching AI copyright challenges.
International Context and Coordination
The Copyright Office has hosted webinars on international copyright issues and artificial intelligence, recognizing that AI technology operates across borders and that international coordination may be necessary for effective policy solutions. The global nature of AI development and deployment means that U.S. copyright policy must consider international approaches and potential harmonization opportunities.
Different countries are taking varying approaches to AI copyright issues, creating potential conflicts and coordination challenges for creators and technology companies operating internationally. The Copyright Office’s attention to international perspectives helps ensure that U.S. policy development considers these broader implications.
Current Status and Future Expectations
As of May 2025, the Copyright Office continues working on the final part of its AI report while implementing the guidance provided in the first two parts. The release of the AI reports has sparked ongoing debate and discussion among stakeholders about the appropriate balance between protecting creator rights and fostering AI innovation.
The office’s work represents an ongoing process rather than a final resolution of AI copyright issues. As AI technology continues advancing and new applications emerge, copyright law and policy will likely need continued adaptation and refinement to address evolving challenges and opportunities.
The Copyright Office’s systematic approach to AI copyright issues, grounded in extensive stakeholder input and careful legal analysis, provides a foundation for addressing current challenges while maintaining flexibility to adapt to future technological developments. The multi-part report structure allows for timely guidance on immediate issues while ensuring comprehensive examination of complex long-term policy questions.
This analysis is based on publicly available information from the U.S. Copyright Office’s published reports and official communications. For the most current information about AI and copyright policy, consult the Copyright Office’s official website and publications.
