Many artists claim that popular generative artificial intelligence services violate copyright law by training them on datasets containing their works, and that in some cases users of these services can directly make copies of their works. Last year, Judge William Orrick allowed a direct copyright infringement complaint to be filed against Stability, the operator of the popular Stable Diffusion AI image generator. But he dismissed various other accusations and asked the artist’s lawyers to make more detailed revisions.
In this recent ruling, revised arguments convinced the judge to grant additional claims that Stability induced copyright infringement. He allowed copyright claims against DeviantArt (which uses a model based on Stable Diffusion) and Runway AI (the original startup behind Stable Diffusion). He also allowed copyright and trademark infringement claims to be brought against Midjourney.
The latter accusation includes accusations that Midjourney misleads users through the “Midjourney Style List,” which includes 4,700 artists whose names can be used to produce works in their style. The artists argued that the list, created without their knowledge or approval, represented a false endorsement, a charge the judge found substantial enough to warrant further argument.
Judge Orrick remained unconvinced by some of the arguments he had previously sent back for more detail. He dismissed claims that generators violated the Digital Millennium Copyright Act by removing or changing copyright management information. He also dismissed claims that DeviantArt violated its terms of service by allowing users’ work to be scraped for use in AI training datasets. Obviously, the claims he allowed will still need to be argued in court.
Kelly McKernan, one of the artists behind the lawsuit, described the ruling as “very exciting” and a “huge victory” for X. Companies in the discovery phase request information, which may reveal details about software tools. “Now we can find out all the things those companies don’t want us to know,” McKernan wrote. (If the company yes Ordered to provide information but not necessarily release it to the public.
But the outcome of the case is difficult to predict. Numerous lawsuits have been filed against artificial intelligence companies, alleging that tools such as Stable Diffusion and ChatGPT make it easy and illegal to copy copyrighted works A lot of training goes into them. The companies counter that the replicas are rare and difficult to produce, arguing that training should be considered a legal fair use. Some earlier lawsuits have been dismissed, including the GitHub Copilot case mentioned in yesterday’s ruling, which was dismissed. Other cases, such as the New York Times Company’s lawsuit against OpenAI, are still ongoing.
Meanwhile, OpenAI, Google, and other tech giants have struck multimillion-dollar deals with publishers (including edge parent company Vox Media) and photo providers for ongoing data access. Smaller companies like Stability and Midjourney have less money to buy access to the data, and individual artists have less leverage to demand payment – so the legal stakes are particularly high for both sides in this dispute.
