OpenAI has been hit with three lawsuits recently, stemming from accusations of violations of copyright laws and extensive data theft.
OpenAI, an American artificial intelligence research company, has become a household name in the past year following the launch of its chatbot, ChatGPT. The chatbot has taken society by storm since its release late last November and claimed 13 million unique users daily in January. The app version of the AI tool quickly became the fastest-growing app of all time (until Meta’s Threads app), amassing 100 million users within three months of its release.
Countless companies have integrated the technology into their business model for assistive purposes as well as for entertainment such as the release of My AI by Snap Inc., the company behind Snapchat.
The program’s capabilities have enabled workers to “work smarter” and increase their efficiency, saving companies significant amounts of capital. The ease of assigning tasks ranging from coding to writing and beyond to the AI tool has resulted in jobs becoming redundant in some industries, with Dropbox, among other companies, citing AI as a reason for cutting its workforce.
In San Francisco on June 28, 2023, two authors filed a class action lawsuit against OpenAI. The lawsuit filed by the authors, according to Andres Guadamuz, a reader in intellectual property law at the University of Sussex, marks the first concerning copyright violations against ChatGPT.
Massachusetts-based writers Paul Tremblay and Mona Awad sued OpenAI in San Francisco federal court claiming, in a proposed class action, that ChatGPT mined data copied from thousands of books without permission, infringing the authors' copyrights https://t.co/OThb8WkvML pic.twitter.com/FcdFWTPFav
— Reuters Legal (@ReutersLegal) June 29, 2023
Paul Tremblay, multiple award-winning and national bestselling author of horror novels such as “The Cabin at the End of the World” and “Growing Things and Other Stories,” and Mona Awad, the author of “Bunny,” (named a Best Book of 2019 by Vogue, Time, and the New York Public Library), and of “13 Ways of Looking at a Fat Girl,” assert that OpenAI violated copyright restrictions through the use of their literature to train the AI tool, ChatGPT.
The authors allege that the summaries generated by the AI tool are “very accurate” and would be impossible to generate without the use of their work in the training of ChatGPT.
While OpenAI has not previously revealed the precise data used to train the chatbot, the company has stated that ChatGPT has absorbed a significant amount of internet data as well as books and Wikipedia.
The fact that books contain “high-quality, well-edited, long-form prose,” makes them ideal for training AI, said the authors’ lawyers, Joseph Saveri and Matthew Butterick, in an email to the Guardian. “It’s the gold standard of idea storage for our species.”
Related Articles: The Race Against AI | AI Could Destroy Humanity in 5-10 Years, 42% of CEOs Say | Who Is Liable if AI Violates Your Human Rights?
The class action lawsuit filed by Tremblay and Awad extends far beyond the writers’ own work. The complaint, according to Reuters, estimated the number of books OpenAI used to train its products at over 300,000, and says this includes copyrighted works from illegal “shadow libraries” offered without permission.
The lawsuit asserts that OpenAI “unfairly” profits from the literary works of the authors and seeks monetary compensation on behalf of all United States-based authors for said damages.
Regardless of whether or not it is true that OpenAI used the works of Tremblay and Awad amongst other authors to train ChatGPT, Guadamuz explains that it will likely be challenging to prove that the authors suffered monetary damages as a result of their literary works being used to train the chatbot.
The second lawsuit was filed against OpenAI on the same day, June 28, claiming that the company secretly stole mass amounts of personal data from the internet to train ChatGPT. The complaint, in almost 160 pages, alleges that the stolen data included “essentially every piece of data exchanged on the internet it could take,” and that OpenAi seized the data without notice, consent or just compensation.”
OpenAI and Microsoft are facing a $3 billion lawsuit over alleged privacy violations related to ChatGPT.
According to the lawsuit, OpenAI secretly "scraped 300 billion words from the internet."
It includes sixteen anonymous plaintiffs.
— TuringPost (@TheTuringPost) July 11, 2023
The complaint filed against OpenAI makes a series of serious accusations regarding OpenAI products and claims that they “use stolen private information, including personally identifiable information, from hundreds of millions of internet users, including children of all ages, without their informed consent or knowledge.”
The lawsuit seeks financial compensation in the form of “data dividends” to those whose personal information was stolen and used to train the products of OpenAI as well as injunctive relief with respect to a short-term block on the commercial use of OpenAI’s products.
The third lawsuit against OpenAI was filed by US comedian and author Sarah Silverman along with two other authors, Christopher Golden and Richard Kadrey, on July 10. Similar to the case presented by Tremblay and Awad, the three authors are suing OpenAI for copyright infringement as they claim the company’s products were trained using their work.
Comedian Sarah Silverman and two authors sued Meta and OpenAI for copyright infringement for allegedly using their content without permission to train artificial-intelligence language models https://t.co/tP23mUaBc0 pic.twitter.com/RkR7QOp40G
— Reuters (@Reuters) July 10, 2023
In the suit, the authors claim that their works were obtained from “shadow library” sites that violate copyright. The authors also included exhibits in the suit exemplifying that, when prompted, OpenAI products provide users with summaries of the authors’ works, including Silverman’s The Bedwetter, Aarat by Golden, and Kadrey’s Sandman Slim.
While Silverman, Golden, and Kadrey present a separate class action lawsuit against OpenAI, the authors are represented by the same lawyers as Tremblay and Awad, Joseph Saveri and Matthew Butterick.
Editor’s Note: The opinions expressed here by the authors are their own, not those of Impakter.com — In the Featured Photo: ChatGPT logo. Featured Photo Credit: Pixabay.