A group of artists filed a lawsuit against three tech companies that created chatbots used to turn text prompts into AI-generated images. The lawsuit is one among many that artists have filed in the past year alleging infringement through use of billions of copyrighted images scraped from the internet. The plaintiff-artists argue that AI chatbots are trained on their works and that these companies don’t provide compensation or obtain consent from the artists before using their work as training material, breaching copyright and intellectual property laws. The defendant-companies creating the chatbots believe that the art used in the training of models is protected under the fair use doctrine and eminent domain, and thus, not subject to copyright infringement. Let’s get into the details.
Background
Last year, three artists filed a US federal class-action lawsuit in San Francisco against AI art companies Midjourney, Stability AI (Stable Diffusion), and DeviantArt. The complaint was filed in January 2023 by three artists representing a larger group of artists “seek[ing] to end [a] blatant and enormous infringement of their rights before their professions are eliminated by a computer program powered entirely by their hard work.”
According to its website, Stable Diffusion is a “deep learning model used for converting text to images. It can generate high-quality, photo-realistic images that look like real photographs by simply imputing any text.” The company’s database houses over 9 million prompts, allowing users to use the free AI image generator to create images. The website also states that “the area of AI-generated images and copyright is complex and will vary from jurisdiction to jurisdiction” and that “images created through Stable Diffusion Online are fully open source, explicitly falling under the CC0 1.0 Universal Public Domain Dedication.”
Over the past year, artists have filed multiple lawsuits against AI tech companies over copyright infringement concerns. According to Benj Edwards writing for Ars Technica: “Since the mainstream emergence of AI image synthesis in the last year, AI-generated artwork has been highly controversial among artists, sparking protests and culture wars on social media.” For more information on the fair use doctrine and artist lawsuits against AI tech, check out our previous blog post on the matter.
What Does the Complaint Allege?
Plaintiffs allege that “[S]tability created and released Stable Diffusion under an open-source license, which DeviantArt and Midjourney use to some extent in their own products.” According to the reporting by FindLaw.com, “to train Stable Diffusion, Stability obtained more than five billion images from an organization, Large-Scale Artificial Intelligence Open Network (LAION), that LAION pulled from the internet.” Though Plaintiffs conceded that the specific images “generated by Stability Diffusion are unlikely to bear a close match to their original content…they claim that because the AI model was trained, in part, on their original content, any images it generates are ‘derivative works’ that infringe their copyrights.”
The complaint put forth seven claims for relief including: direct copyright infringement, violation of the Digital Millennium Copyright Act, violation of the common law right of publicity, and unfair competition to name a few.
What has the Court Decided so far?
In November 2023, the federal judge presiding over the case, Honorable Judge William Orrick, streamlined the class action by “dismissing most claims without prejudice (meaning Plaintiffs may replead their claims),” but allowed “the core claim of copyright infringement against Stability AI [to] stand for now.”
The judge’s decision started with dismissing copyright infringement claims “brought by artists who failed to register their copyrights with the Copyright Office” and “limited the claims brought by a third artist to only those images she had in fact registered.” The judge did side with Plaintiffs regarding their AI training claim and found “that the artist had plausibly alleged that her entire collection had been scrubbed from the internet and included in the LAION dataset.”
All other claims asserted by Plaintiffs, “including vicarious infringement and unfair competition, among others,” were dismissed by the judge, “but he gave the Plaintiffs leave to replead them in an amended complaint, and essentially provided them with an outline of how the Plaintiffs could avoid dismissal in the future.”
Broader Implications – Fair Use or Copyright Infringement
The case at hand, Andersen v. Stability AI, Ltd., stands as a reminder that content creators concerned about copyright infringement must first register their copyrights to be eligible to sue AI companies for training infringement. An important decision concerning AI training models as “it appears to be the first in the country in which a federal judge has suggested that you can’t train AI on someone’s copyrighted work without their permission.”
Large AI language models (LLMs) are a relatively new technology, and “courts have never ruled on [the] copyright implications.” Some experts believe “copyright’s fair use doctrine allows Stability AI to use” images scraped from the internet to train their chatbots. Others argue, that “[t]here’s a real possibility that the courts could decide that Stability AI violated copyright law on a massive scale.”
The court’s decision will greatly impact AI tech and legal communities. If the court rules in favor of Plaintiffs, “building cutting-edge generative AI would require licenses from thousands –perhaps even millions—of copyright holders.” A process likely to be costly and time-consuming, meaning only a handful of the largest tech companies could afford to do it –gaining an edge over smaller tech AI start-ups. Further, such a restriction could mean the “resulting models likely wouldn’t be as good. And smaller companies might be locked out of the industry altogether.”
If, however, the court rules in favor of the tech companies training AI models, the impact of such a decision could have devastating and lasting consequences for communities of creators –from writers to artists, musicians, and authors threatened by AI art generators, many of whom percieve the technology as a risk to their livelihood. AI art generators create content through image synthesis, and artists are concerned that these tech companies have infringed on their “intellectual property to train their models” without consent from the artists and without providing compensation for using their work. For more information on AI-related lawsuits, check out our past blog posts and check back in with us as we continue to track these lawsuits that could have significant ramifications for the fair use doctrine in copyright law.
Interested in Learning More?
Check out Trellis! Trellis is an AI-driven, state trial court research and analytics platform. We make the fragmented U.S. state trial court system searchable through a single interface, offering comprehensive insights into judges, cases, and opposing counsel. Effortlessly track lawsuits across states and stay up-to-date with ongoing litigation documents. Request a demo today and elevate your legal practice with our intuitive analytics and API.
Sources:
https://stablediffusionweb.com/#faq
Music: Disruptor’s Dance by Anka Mason
Blog Narration: Anka Mason
