U.S. Court Orders Artists to File a New Complaint After Finding Flaws in Lawsuit Against AI Companies
The U.S. District Judge William Orrick after noting flaws in the lawsuit against artificial intelligence (AI) companies, stated that he was inclined to dismiss most of the lawsuit brought by a group of artists against generative artificial intelligence companies. However, he allowed them the opportunity to file a new complaint. The artists accused Stability AI, Midjourney, and DeviantArt of misusing their work to train their AI systems.
The Judge noted that the artists should be able to provide more factual evidence about the alleged copyright infringement as they have access to Stability AI’s relevant source code. The Court suggested that without additional evidence, it would be implausible to assume that the plaintiffs’ works were involved in the training of the AI systems, which have been trained on five billion compressed images.
In January the plaintiffs, including Sarah Andersen, Kelly McKernan, and Karla Ortiz, alleged in their complaint filed, that Stability AI scraped billions of images from the internet without permission to train their Stable Diffusion text-to-image system.
However, the Judge also mentioned that illustrator Sarah Andersen’s claim could potentially survive the initial bid to dismiss the lawsuit. Andersen claimed that Stability AI directly infringed the copyrights she had registered for her works.
They accused the company of copyright infringement by using their work to create AI-generated images in their styles. Midjourney and DeviantArt, whose AI systems incorporate the Stable Diffusion technology, were also named as defendants in the lawsuit.
Judge Orrick clarified that it was unclear whether the artists were accusing Midjourney and DeviantArt of copyright infringement through their implementation of Stability’s model or by training their own systems in an infringing manner.
The Court further noted that the artists were unlikely to succeed in claiming that the images generated by the AI systems, based on text prompts using their names, violated their copyrights as there was insufficient similarity between the artists’ work and the generated output.
Orrick asserted that there was currently no substantial similarity between the artists’ created images and those generated by the AI systems.
“I don't think the claim regarding output images is plausible at the moment, because there's no substantial similarity between images created by the artists and the AI systems”, Orrick said.
Therefore, on recognizing the flaws in the artists’ lawsuit against generative AI companies, the Judge expressed his inclination to dismiss most of their claims. However, he allowed the artists to file a new complaint to address the issues with clearer and more differentiated claims against the defendants.
Notably, several lawsuits have been filed against companies like Microsoft, Meta, and OpenAI in the rapidly growing generative AI field.