For years now, the intersection of artificial intelligence and copyright law has been the center of attention for artists, creators, software developers, entrepreneurs, technologists, and of course, lawyers. Back in 2023, we wrote about the U.S. Copyright Office’s decision to register certain components of Zarya of the Dawn, a comic book featuring artwork generated by the artificial intelligence tool Midjourney. (In case you need a refresher, the Copyright Office canceled the author’s existing registration for the comic book as a whole due to failure to exclude AI-generated images, and replaced it with a registration covering only the artist’s original text and the selection, coordination, and arrangement of that text with the AI-generated images—specifically excluding the images themselves. The Copyright Office clarified that its decision was based on the way in which Midjourney operates, and that creators using “other AI offerings” may see a different result and become the author of AI-generated images for copyright purposes.)

Since that time, generative AI technology has continued to improve and proliferate, raising new questions about when the men and women behind the machines can claim copyright ownership of the generated outputs, including text, visuals, and sounds.

Meanwhile, the Copyright Office has requested, received, read, and considered over 10,000 comments from “stakeholders” across various industries, including “authors and composers, performers and artists, publishers and producers, lawyers and academics, technology companies, libraries, sports leagues, trade groups and public interest organizations, and even a class of middle school students.” The comments were originally submitted in response to the Office’s 2023 Notice of Inquiry on Copyright and AI. About half them specifically addressed the copyrightability of AI-generated outputs.

In January of 2025, the Copyright Office issued its highly-anticipated report on the copyrightability of AI-generated outputs. So, are they copyrightable? The answer, unsurprisingly, is that it depends “on the nature and extent of a human’s contribution, and whether it qualifies as authorship of expressive elements contained in the output”. 

To provide further clarity on the issue of copyrightability and AI outputs, the Office ultimately made the following conclusions and recommendations, laid out in the Executive Summary of the report:

1. Questions of copyrightability and AI can be resolved pursuant to existing law, without the need for legislative change. 
2. The use of AI tools to assist rather than stand in for human creativity does not affect the availability of copyright protection for the output. 
3. Copyright protects the original expression in a work created by a human author, even if the work also includes AI-generated material. 
4. Copyright does not extend to purely AI-generated material, or material where there is insufficient human control over the expressive elements. 
5. Whether human contributions to AI-generated outputs are sufficient to constitute authorship must be analyzed on a case-by-case basis. 
6. Based on the functioning of current generally available technology, prompts do not alone provide sufficient control. 
7. Human authors are entitled to copyright in their works of authorship that are perceptible in AI-generated outputs, as well as the creative selection, coordination, or arrangement of material in the outputs, or creative modifications of the outputs. 
8. The case has not been made for additional copyright or sui generis protection for AI generated content.

Further Clarification from the Copyright Office: Control is Key

At first glance, it may seem like not much has changed from 2023, and not much will change any time soon when it comes to copyright and AI. Importantly, however, the Office has clarified a number of issues in its report for anyone using generative or other AI tools as part of their creative processes.

Ultimately, if a human does not have sufficient control over the AI-generated output, that output is not eligible for copyright protection. As was the case was in 1965, when the Office wrestled with questions of authorship raised by developments in computer technology, the “crucial question” when it comes to copyrightability of AI-generated output is “whether the “work” is basically one of human authorship, with the computer merely being an assisting instrument, or whether the traditional elements of authorship in the work (literary, artistic, or musical expression or elements of selection, arrangement, etc.) were actually conceived and executed not by man but by a machine.”

In the 2025 report, the Copyright Office carefully distinguished between the different ways humans can use modern AI tools to generate new works and considered the copyright implications of each. These include: (1) use of AI as an “assistive instrument,” (2) prompts, (3) expressive inputs, and (4) modifications made to, and selection and arrangement of, AI-generated content.

Assistive Use

First, the Office clarified that “assistive uses of AI systems” should not affect the availability of copyright protection for the output.  In other words, using AI in the creative process is not automatic grounds for disqualification when it comes to copyright protection. Moreover, adding AI-generated output to a larger creative work does not mean the larger work is ineligible for copyright protection either.

The Office specifically pointed to various tasks that have been performed in creative fields for years, some of which are now done using AI: “aging” or “de-aging” actors, identifying chord progressions, detecting errors in software code, and removing unwanted objects or crowds from a scene, to name a few. According to the Copyright Office, because these and similar uses of AI do not actually stand in for human creativity, they should not limit copyright protection for the final works. The result may be different where an AI tool “makes expressive choices.”

Prompts

The Copyright Office concluded, in short, that prompts alone do not give anyone sufficient control over AI-generated outputs to make them the “author” of those outputs for copyright purposes, at least with today’s technology.  In the Office’s view, prompts are essentially just “instructions that convey unprotectible ideas.”

Interestingly, in 2023, the Copyright Office explained for Zarya that the human prompter behind the AI output from Midjourney could not be the “author” of the AI-generated images, despite writing “thousands” of descriptive text prompts, because Midjourney generates images using those prompts in an “unpredictable way.” Addressing the same technology in the 2025 report, the Office clarifies that when it comes to whether a human being can be considered the “author” of a work generated by AI (which is required for copyright protection), the question is “the degree of human control, rather than the predictability of the outcome.”

The Office explained in its report that with most generative AI tools, “gaps between prompts and resulting outputs demonstrate that the user lacks control over the conversion of their ideas into fixed expression, and the system 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 Office illustrates this point with the following prompt and resulting output from Midjourney:

Here, the AI-generated image reflects some of the instructions provided in the Office’s written prompt (e.g., “a bespectacled cat smoking a pipe”), but not others (e.g., “a highly detailed wood environment”). Notably, the AI system filled in the gaps where specific instructions were missing. Among other things, the AI system determined the cat’s breed, coloring, size, pose, facial expression, and the fact that it has a human hand instead of a paw. Again, these decisions by the AI system show that the human behind the prompt does not sufficiently control the output to warrant copyright protection.

As for meticulous and prodigious prompters, the Office took the position that an artist who “repeatedly enters prompts until the output matches their desired expression” is essentially spinning a metaphorical wheel with infinite possibilities. It concluded that someone doing so and then selecting from the options presented is insufficient to claim ownership of the resulting outputs. “The fact that identical prompts can generate multiple different outputs further indicates a lack of human control.” 

Finally, the Copyright Office noted that advances in technology can either provide users with increased control over the expressive elements of AI-generated outputs or, alternatively, limit that control for the sake of automation or optimization. As the degree of creative control over outputs changes, the copyrightability of those outputs may change as well.

Expressive Inputs

“Expressive inputs” refer to inputs in the form of text, images, audio, and/or video that can be “substantially retained as part of the output” (which may occur when using specific tools, settings, or prompts). For example, an artist may create an original visual work, input that work into an AI system, and instruct the system to change the colors or layering of objects. Or, a writer may upload an original story written in first person and instruct the AI system to convert it to a third-person narrative. 

The Office concluded that “where 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.” More specifically, “their own creative expression will be protected by copyright, with a scope analogous to that in a derivative work. Copyright in this type of AI-generated output would cover the perceptible human expression. It may also cover the selection, coordination, and arrangement of the human-authored and AI-generated material, even though it would not extend to the AI generated elements standing alone.”

To demonstrate this practically, the Copyright Office uses the following written prompt, expressive input, and output:

Here, the author created an expressive image (input) by hand and entered it into an AI system with a written prompt in order to create a new image (output).

The Office noted that here, “the drawing itself is a copyrightable work, and its expressive elements are clearly perceptible in the output, including the outline of the mask, the position of the nose, mouth, and cheekbones relative to the shape of the mask, the arrangement of the stems and rosebuds, and the shape and placement of the four leaves.”

Importantly, the copyright applicant in this work disclaimed “any non-human expression” appearing in the final work, such as “the realistic, three-dimensional representation of the nose, lips, and rosebuds, as well as the lighting and shadows in the background.” 

Ultimately, the office registered the output above, noting that registration is “limited to unaltered human pictorial authorship that is clearly perceptible in the deposit and separable from the non-human expression that is excluded from the claim.” 

Modifying or Arranging AI-Generated Content

The Copyright Office reaffirmed that it’s possible to select, coordinate, and arrange AI-generated material in a sufficiently creative way to warrant copyright protection in the resulting work as an original work of authorship. However, the wholly-AI generated components are not protectable individually.

Citing its 2023 decision regarding Zarya of the Dawn as an example, the Office explained that comic book is “the product of creative choices with respect to the selection of the images that make up the Work and the placement and arrangement of the images and text on each of the Work’s pages. Copyright therefore protects [the applicant’s] authorship of the overall selection, coordination, and arrangement of the text and visual elements that make up the Work.” 

The Office also reaffirmed that one can modify, edit, or adapt AI-generated outputs to such a degree that the modifications meet the standard for copyright protection. Similar to the way copyright applies to derivative works, the copyright in the edited output would “extend to the material the human author contributed but would not extend to the underlying AI-generated content itself.”

Still using Midjourney as an example, the Office pointed out that certain AI tools (like Midjourney’s “Vary Region and Remix Prompting”) allow users to edit AI outputs “in an iterative fashion” by re-generating chosen elements within the outputs using written prompts.

The Office noted that such tools give users a much greater ability to “control the selection and placement of individual creative elements” in output more than just written prompts alone. Those outputs may be copyrightable when the modifications meet the minimum standards of originality for protection, as determined on a case by case basis.

Final Thoughts

AI hasn’t turned copyright law on its head. At least not yet. For now, copyright will remain flexible and respond and adapt to AI and other new technologies, just as it’s done before with the introduction of photographs, motion pictures, computer programs, and video games.

By clarifying that the critical question when it comes to copyrightability of AI outputs is the degree of human control over those outputs, and further analyzing the amount of control when using different methods to generate outputs available today, the U.S. Copyright Office’s 2025 report on copyrightability and AI offers greater clarity for creators, businesses, and legal professionals. 

Of course, as generative AI becomes more sophisticated and commonplace, new questions about authorship, ownership, and creative control will arise. We will keep you updated as these questions are asked and answered.