Ai-Created Book Wins Copyright, but There’s a Catch
Elisa Shupe Wins Copyright Case with AI-Generated Book
Elisa Shupe, a retired U.S. Army Veteran, has won “round two” with the U.S. Copyright Office, obtaining copyright registration of her self-published book, which was created using ChatGPT, an artificial intelligence application. While she lost round one – her original application was initially rejected as machine-generated elements (including AI-generated components) generally are excluded from copyright protection – she ultimately won on appeal.
AI Machinations: Shupe’s Unique Autofiction Book
Shupe’s self-published autofiction book AI Machinations: Tangled Webs and Typed Words, which is available on Amazon under the pseudonym Ellen Rae, details Shupe’s eventful life, including her fight for broader gender recognition. Shupe’s work is among the first creative works to obtain copyright registration for the arrangement of AI-generated text.
In her original application, Shupe asserted an ADA exemption for her “many disabilities” including bipolar disorder, borderline personality disorder, and brain stem malformation, claiming the use of AI-generated text served as assistive technology to aid her in communicating, similar to an amputee utilizing a prosthetic leg. Shupe, who is categorized as 100 percent disabled by the U.S. Dept. of Veteran Affairs, claims these conditions cause cognitive impairment and that she would not have been able to complete the book without the use of ChatGPT.
Legal Assistance and the Appeal Process
When the original application was rejected, Shupe sought assistance from Jonathan Askin, founder of the Brooklyn Law Incubator & Policy Clinic at Brooklyn Law School, which focuses on pro bono representation in cases involving emerging technology and policy matters. Assisted by law student Sofia Vescovo, Askin crafted Shupe’s appeal to the U.S. Copyright Office, not only expounding on her claim of ADA exemption, and a claim that the USCO had discriminated against her based on her disabilities, but also detailing the extensive work she performed in compiling the book.
The appeal provided a lengthy account of how Shupe revised and re-arranged the ChatGPT-generated text, which she says involved 14-hour workdays, as well as her custom commands and edits. The submission provided a side-by-side presentation of the original ChatGPT-generated text and Shupe’s final version and cited to the copyright registration issued in February 2023 for Zarya and the Dawn, a graphic novel that includes AI-generated images.
Copyright Decision: A Limited Victory
In awarding the copyright registration, the USCO did not speak to the disability arguments, but concluded Shupe could be awarded copyright protections for the “selection, coordinating, and arranging of text generated by artificial intelligence.”
This means that while Shupe is not considered the author of the entire work, the book itself cannot be copied, although, at least in theory, the individual sentences could be rewritten, re-arranged, and published as a different book. Shupe maintains that her use of ChatGPT more closely resembles a collaboration than automated output but considers the limited copyright registration a win “for now.”
Uncertain Future for Copyright and AI
The full ramification of this registration remains unclear but raises any number of uncertainties for the future. Given that the USCO did not address the disability arguments in Shupe’s original application and subsequent appeal, it is unclear whether those arguments could be used to support later submissions from persons using some form of generative-AI.
Still further, given the significant amounts of information provided by Shupe regarding her arrangement and editing process, what amount of revision is necessary to support even a limited copyright registration? In an executive order issued last fall, President Biden asked the U.S. Copyright Office and U.S. Patent and Trademark Office to consult and make recommendations on AI to the White House, including on the “scope of protection for works produced using AI.” Yet, it remains unclear where these offices will draw the line.
Conclusion: The Growing Role of AI in Copyright
Finally, what impact, if any, will these types of registrations have on the many pending lawsuits surrounding the use of copyrighted content to train AI bots (such as ChatGPT) without authorization or licensing? Will even a limited copyright registration (and the ability to purchase this content on Amazon) bolster the claim for damages by the New York Times, Authors Guild, and others in pending copyright infringement lawsuits? The landscape certainly remains rocky as AI-proponents, content generators, and consumers navigate the Wild West of generative artificial intelligence. And whether Congress – or even the individual states – will enact legislation specific to these concerns is likewise unclear. What remains certain is the ever-growing use of generative AI, which now includes copyright protection, at least to a limited degree.
Klemchuk PLLC is a leading IP law firm based in Dallas, Texas, focusing on litigation, anti-counterfeiting, trademarks, patents, and business law. Our experienced attorneys assist clients in safeguarding innovation and expanding market share through strategic investments in intellectual property.
This article is provided for informational purposes only and does not constitute legal advice. For guidance on specific legal matters under federal, state, or local laws, please consult with our IP Lawyers.
© 2024 Klemchuk PLLC | Explore our services