No, ChatGPT did not Write this Article – But if it Did, Could it be Copyrighted?

Reni de la NuezNews & Insights

By Reni de la Nuez

In September, we published an article entitled “Looking Beyond the FN Meka Scandal: AI & Copyright.”[1] This article examined emerging uncertainty surrounding the copyrightability of artificial intelligence-generated works, given that the US Copyright Office (hereinafter the “Office”) has consistently held that human authorship is necessary in order for a work to be copyrighted.[2] In the time since that article was published, the Office has determined that the increasing prevalence of AI tools in creative works indicates the need for a more flexible approach to addressing AI as it relates to copyright law.[3]

On March 16, 2023, the Office released a statement regarding its approach to AI considerations in copyright applications entitled “Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence.” The statement clarified the Copyright Office’s practices for examining and registering AI works.[4] For example, the Copyright Office does not, at least at the present moment, intend to assign copyrights to any non-human author.[5] Citing the 1884 case Burrow-Giles Lithographic Co. v. Sarony,[6] the Office notes that courts have historically considered the authorship requirement of copyright to be satisfied only by a human author who is entitled to “the exclusive right . . . to the production of [their] own genius or intellect.”[7] However, this seemingly simple principle is complicated by the increasing popularity and accessibility of AI software.[8]

In order to better understand the scope of this issue, it is important to consider the way that the AI tools function. Generally speaking, the human user will provide a written prompt that directs the AI system to produce a written text, image, or sound bite in response.[9] In order to generate the final product, the AI system will rely on the existing materials to which it has access and make an original work in response to the prompt.[10] However, the Office has explained that “[w]hen an AI technology determines the expressive elements of its output, the generated material is not the product of human authorship. As a result, that material is not protected by copyright and must be disclaimed in a registration application.”[11] To illustrate, the Office referenced a Ninth Circuit case in which the court determined that a monkey who took a “selfie” photo did not have standing to sue under the Copyright Act.[12]

A recent example that demonstrates the issues the Copyright Office is currently grappling with involved the comic book “Zarya of the Dawn” in which the author had written the text and Midjourney’s generative AI technology had produced the accompanying artwork.[13] The author failed to disclose this separation of authorship in their application for the copyright.[14] Upon discovering this fact, the Copyright Office solicited information from the author as to why the registration should not be canceled and ultimately canceled the approval that they initially gave the author and replaced it with a more limited registration.[15] Instead of allowing the author to hold the copyright to the work in its entirety, the Office stated that the human creator was “the author of the Work’s texts as well as the selection, coordination, and arrangement of the Work’s written and visual elements” but could claim no right to the images that were created by AI technology, thereby rendering them unprotected under current copyright law.[16] In consideration of cases like this, the Office’s recent statement provides that, “applicants have a duty to disclose the inclusion of AI-generated content in a work submitted for registration and to provide a brief explanation of the human author’s contributions to the work.”[17]

The statement additionally details some current examples for determining whether an author’s use of AI as a creative tool might push the work into the uncopyrightable category, but it is unclear whether the current guidelines will apply long-term.[18] On March 23rd, at The Copyright Society’s annual “Copyright and the California Coast” meeting, Director of the U.S. Copyright Office Shira Perlmutter discussed the lack of rules in this area of law and stated that the Copyright Office plans to review each application in consideration of the amount of AI elements included and the way in which the specific AI tool that was used functions.[19] She reiterated that the office presently rejects the establishment of any hard parameters and will instead opt for a fact-specific analysis of each application. So long as the Office determines that the human author engaged in the requisite minimum amount of originality for a given element of the work, that element – but not necessarily the entire product – will be copyrightable.[20] The Office’s determinations are based solely on the facts disclosed by the applicant, and the Office will not further investigate whether the application truthfully depicts the creative process.[21]

The Office’s lack of specificity has generated disagreement from those both for and against the copyrightability of AI works.[22] Those in favor of copyrighting AI works, such as the lawyers who represented the “Zarya of the Dawn” author, argue that the human’s creative use of the AI system should be given as much weight as the AI output.[23] Those opposed have expressed concern with the threats that this would pose to the existing protected works that are potentially involved in creating an AI work.[24]

Since this is still a developing area, the Office has dates planned in April and May for in-depth discussions regarding specific copyright issues with a “focus on literary works, including print journalism and software; visual arts; audiovisual works; and music and sound recordings.”[25] Participants will be asked to “discuss their hopes, concerns, and questions about generative AI and copyright law.”[26] Links to these virtual meetings can be found here and recordings of the events will also be posted for public view.

[1] Reni de la Nuez, Looking Beyond the FN Meka Scandal: AI & Copyright, Loyola Law School Legal Developments (Sept. 30, 2022),

[2] Id.

[3] Copyright Office Launches New Artificial Intelligence Initiative, U.S. Copyright Office NewsNet Issue 1004 (Mar. 16, 2023),

[4] “Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence,” 88 Fed. Reg. 16190 (Mar. 16, 2023) (to be codified at 37 C.F.R. 202),

[5] Id. at 16191.

[6] Burrow-Giles Lithographic Co. v. Sarony, 111 U.S. 53, 56 (1884).

[7] Id.  

[8] Id.

[9] Id. at 16192.

[10] Id.

[11] Id.

[12] Naruto v. Slater, 888 F.3d 418, 426 (9th Cir. 2018), decided on other grounds.

[13] Letter from U.S. Copyright Office to Van Lindberg regarding Zarya of the Dawn, Registration No. VAu001480196 (Feb. 21, 2023),

[14] Id. at 2.

[15] Id. at 1-3, 12.

[16] Id. at 1, 4-5.

[17] “Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence,” 88 Fed. Reg. 16190, 16193 (Mar. 16, 2023) (to be codified at 37 C.F.R. 202),

[18] Id. at 16192.

[19] Copyright and the California Coast: The Register of Copyrights’ Address (Mar. 23, 2023). 

[20] Feist Publ’ns, Inc. v. Rural Tel. Serv. Co, 499. U.S. 340, 348 (1991) (“The mere fact that a work is copyrighted does not mean that every element of the work may be protected. Originality remains the sine qua non of copyright; accordingly, copyright protection may extend only to those components of a work that are original to the author.”).

[21] Copyright and the California Coast: The Register of Copyrights’ Address (Mar. 23, 2023).  

[22]  Tony Analla, Zarya of the Dawn: How AI is Changing the Landscape of Copyright Protection, Harvard Law School Jolt Digest (Mar. 6, 2023),

[23] Id.

[24] Id.

[25] U.S. Copyright Office, Copyright and Artificial Intelligence, Spring 2023 AI Listening Sessions,

[26] Id.