Copyright Office: No Sui Generis Protection for AI-Generated Material

On January 29, 2025, the Copyright Office published Part 2 of its multipart series of reports on copyright law and artificial intelligence examining copyrightability (Report). The Report summarized over ten thousand comments received in response to the Copyright Office’s August 2023 Notice of Inquiry, and emphasized long-standing principles that only works of human authorship are eligible for copyright protection, which is determined on a case-by-case basis. The Report concluded that copyright does not extend to material purely generated by artificial intelligence, or AI; prompts alone are insufficient as human authorship; copyrightability of works involving the use of AI can be resolved under existing law without the need for legislative change; and there is presently no need for sui generis protection for AI-generated material.

A Brief History of Technology and Copyright

Copyright law in the United States is an incentive-based system. The Copyright Act provides certain exclusive rights as economic incentives to create copyrightable works. The purpose for providing economic incentives is to ultimately increase the amount of material in the public domain. The Second Circuit has described the public domain as “the common wellspring.” “In this pool are not only elemental ‘raw materials,’ like colors, letters, descriptive facts, and the catalogue of standard geometric forms, but also earlier works of art that, due to the passage of time or for other reasons, are no longer copyright protected.” Thus, the true purpose of the Copyright Act is to ensure a steady flow of works into the public domain, and providing economic incentives to create copyrightable works is a means to achieve that purpose.

Copyright law has always followed advancements in technology but has been flexible enough to adapt to those advancements. The US Supreme Court observed, “[f]rom its beginning, the law of Copyright has developed in response to significant changes in technology. Indeed, it was the invention of a new form of copyright equipment—the printing press—that gave rise to the original need for copyright protection.” However, human input has always remained a constant for copyrightability, despite advances in technology. “Copyright is designed to adapt with the times. Underlying that adaptability, however, has been a consistent understanding that human creativity is the sine qua non of copyrightability, even as that human creativity is channeled through new tools or into new media.”

It is against this backdrop of the Copyright Act’s history of adaptability that the Report explained how human-authored expression has always been required. The Report described how in 1884, Burrow-Giles Lithographic Co. v. Sarony extended copyright protection to photographs because, despite being a product of new technology—the camera—human expression was present in the photographs the camera produced. The Burrow-Giles Court recognized the requirement of human authorship by describing copyright as “the exclusive right of a man to the production of his own genius or intellect.”

The Report also described how ideas and facts are not protectible by copyright law, and how the US Supreme Court rejected the “sweat of the brow” doctrine in Feist Publications, Inc. v. Rural Telephone Service Co. The Report observed, “[n]o court has recognized copyright in material created by non-humans, and those that have spoken on this issue have rejected the possibility.” Concluding the backdrop of the Copyright Act’s adaptability and how facts, ideas, and effort are not protectible, the Report stated, “[i]n sum, the use of a machine as a tool does not negate copyright protection, but the resulting work is copyrightable only if it contains sufficient human authored expressive elements.” The Report then analyzed three kinds of human contribution to AI-generated material: “(1) prompts that instruct an AI system to generate an output; (2) expressive inputs that can be perceived in AI-generated outputs; and (3) modifications or arrangements of AI-generated outputs.”

Copyrightability of Prompts

The Report next analyzed the copyrightability of prompts. It opined that “there is an important distinction between using AI as a tool to assist in the creation of works and using AI as a stand-in for human creativity.” It restated commenters’ concerns of AI systems as unpredictable “black boxes” and highlighted comments about how prompting “can require a significant amount of time and demonstrable human effort” to come up with an output they like, echoing the “sweat of the brow” doctrine that Feist retired over thirty years ago.

Other commenters “likened repetitive prompting to ‘spinning a roulette wheel with infinite possibilities. [ ] When a user [metaphorically] spins the wheel dozens of times until they land on an output they like,’ such activity should not give the user a right to claim ownership of that output.” The Copyright Office reaffirmed the irrelevancy of the “sweat of the brow” doctrine as applied to prompts, holding:

[T]he time, expense, or effort in creating a work by revising prompts is irrelevant, as copyright protects original authorship, not hard work or “sweat of the brow.” [ ] By revising and submitting prompts multiple times, the user is “re-rolling” the dice, causing the system to generate more outputs from which to select, but not altering the degree of control over the process.

The Copyright Office found that the “black box” gaps between prompts and resulting outputs demonstrated that the user lacked control over the conversion of their ideas into fixed expression, and it is the AI system, not the user, that is largely responsible for determining the expressive elements in the output. “In other words, prompts may reflect a user’s mental conception or idea, but they do not control the way that idea is expressed.” The Copyright Office concluded:

[G]iven current generally available technology, prompts alone do not provide sufficient human control to make users of an AI system the authors of the output. Prompts essentially function as instructions that convey unprotectable ideas. While highly detailed prompts could contain the user’s desired expressive elements, at present they do not control how the AI system processes them in generating the output.

Copyrightability of Expressive Inputs Perceivable in AI-Generated Outputs

The Report analyzed the copyrightability of expressive inputs where human expression was discernable in the AI-generated output. An example of an expressive input is where a human provides a hand-drawn image as an input, and the AI system adds to or modifies the image. The Copyright Office highlighted commenters who opined that where human-authored inputs are reflected in the output, the human author may have a greater claim to authorship than mere prompts alone. The Report concluded:

[W]here a human inputs their own copyrightable work and that work is perceptible in the output, they will be the author of at least that portion of the output. Their own creative expression will be protected by copyright, with a scope analogous to that in a derivative work. Just as a derivative work protection is limited to the material added by the later author, copyright in this type of AI-generated output would cover the perceptible human expression [ ] though it would not extend to the AI-generated elements standing alone.

Copyrightability of Modifications or Arrangements of AI-Generated Outputs

The first two analyses above focused on human inputs and the resulting output. This third analysis focuses on human modification of AI-generated material, analogous to the copyrightability of compilations under current copyright law. “For example, a film that includes AI-generated special effects or background artwork is copyrightable, even if the AI effects and artwork separately are not.”

The Report highlighted comments that opined, “human authors should be able to claim copyright if they select, coordinate, and arrange AI-generated material in a creative way” and modifications “should be assessed in the same way as . . . editorial or other changes to a pre-existing work.” It concluded that human modifications or arrangements of AI-generated outputs could be copyrightable, but such determination would need to be made on a case-by-case basis. “Again, the copyright would extend to the material the human author contributed but would not extend to the underlying AI-generated content itself.”

No Sui Generis Protection for AI-Generated Material

In addition to analyzing the copyrightability of AI-generated material, the Report also examined the current incentive-based framework if applied to such material. Many comments expressed concern that if AI-generated material were entitled to copyright protection, it would result in a flood of such material, stifle human creativity, and reduce human-authored works, “crowding and diluting the marketplace to the point that copyright incentives no longer function as intended.” Others predicted that incentivizing AI-generated material would disincentivize human-authored works because “the cost of human creation would be deemed too burdensome in comparison to using AI.”

Most commenters were against sui generis protection for AI-generated material. The Copyright Office agreed, explaining:

In the Office’s view, the case has not been made for additional protection for AI-generated material beyond that provided by existing law. … [I]t is not clear that new incentives are needed. The developers of AI models and systems already enjoy meaningful incentives under existing law (as indicated by the rapid development and adoption of those models and systems). These incentives include patent, copyright, and trade-secret protection[.] Moreover, we are not convinced that providing further incentives would promote progress. We share the concerns expressed about the impact of AI-generated material on human authors and the value that their creative expression provides to society. If a flood of easily and rapidly AI-generated content drowns out human-authored works in the marketplace, additional legal protection would undermine rather than advance the goals of the copyright system. The availability of vastly more works to choose from could actually make it harder to find inspiring or enlightening content. Indeed, AI training itself is reportedly reliant on human-generated content, with synthetic data leading to lower-quality results.

In concluding that no sui generis protection is needed for AI-generated material, the Report noted that an estimated 170 million AI-generated music tracks already threaten to dilute human creators’ royalties and observed: “if authors cannot make a living from their craft, they are likely to produce fewer works. And in our view, society would be poorer if the sparks of human creativity become fewer or dimmer.”

AI as a Tool for Disabled Creators

The Report addressed whether protection for AI-generated material would empower individuals with disabilities to create. It acknowledged the potential of certain technologies, such as image generation and text-to-speech, to improve communication of individuals with certain disabilities. The Copyright Office described how AI technology helped GRAMMY-winning country artist Randy Travis record a song, “Where That Came From,” after his speech was impaired following a stroke. The Copyright Office explained that “[b]ecause the sound recording used AI as a tool, not to generate expression, the Office registered the work.” Implying that existing law adequately provides for AI to be used as a tool for creators who have disabilities, the Copyright Office stated:

The Office strongly supports the empowerment of all creators, including those with disabilities. We stress that to the extent these functionalities are used as tools to recast, transform, or adapt an author’s expression, copyright protection would be available for the resulting work.”

In Conclusion: No New Legislation Recommended

The Report concluded by acknowledging the public’s desire for clarity around the copyrightability of AI-generated material but opined that “existing legal doctrines are adequate and appropriate to resolve questions of copyrightability.” The Report maintained that the copyrightability of AI-generated material must be made on a case-by-case basis, but “[p]rompts alone, however, at this stage are unlikely to satisfy those requirements.” Because copyrightability is determined on a case-by-case basis, “statutory language would be limited in its ability to provide brighter lines. Unless and until future developments raise new problems, the Office does not recommend a change in the law.”