A group of artists who sued the makers of an artificial intelligence (AI) software that generates images from text prompts have lost most of their claims in a federal court. The artists accused the companies of using their work without permission or compensation to train the AI tool, and of creating unfair competition in the market.
The lawsuit was filed in January by Sarah Andersen, Kelly McKernan, and Karla Ortiz against Stability AI, Midjourney, and DeviantArt. The plaintiffs are professional artists who create original illustrations, paintings, and comics. They claimed that the defendants used billions of copyrighted images, including theirs, to train Stable Diffusion, an AI tool that can produce images based on text inputs.
The artists argued that Stable Diffusion infringed their rights by copying their styles and creating similar images that could replace their work. They also alleged that the defendants violated their right of publicity by using their names as keywords to generate images in their style. They sought a permanent injunction to stop the use of their work and damages for their losses.
However, on Monday, Senior U.S. District Judge William Orrick dismissed most of the claims, finding them too vague or unsupported by facts. He only allowed one claim of direct infringement to proceed against Stability AI, which allegedly downloaded or acquired the images without authorization.
Judge Orrick ruled that the other claims failed to show that the images generated by Stable Diffusion were substantially similar to the plaintiffs' work, which is a key requirement for infringement. He also found that the plaintiffs did not adequately allege how the defendants used their names or how that harmed them.
The judge also rejected the plaintiffs' argument that Stable Diffusion violated the Digital Millennium Copyright Act (DMCA) by circumventing their technical measures to protect their work. He said that the plaintiffs did not specify what those measures were or how they were bypassed.
The defendants have argued that their use of the images was protected by fair use, a legal doctrine that allows limited use of copyrighted material for purposes such as criticism, comment, news reporting, teaching, scholarship, or research.
Rising Number of Lawsuits Over AI Copyrights
The case is one of the first to challenge the legality and ethics of AI image generators, which have become increasingly popular and accessible in recent years. The technology has raised questions about the boundaries of creativity, originality, and ownership in the digital age.
In June, partners Microsoft and OpenAI were taken to court over how they train their AI models. The lawsuit claims that OpenAI secretly collected 300 billion words from the internet, such as books, articles, websites and posts with personal information that was not given willingly. The plaintiffs say that OpenAI broke privacy laws, terms of service contracts, and computer fraud laws by using their private data without their consent or awareness.
Earlier this month, Universal Music Group confirmed it is suing AI developer Anthropic. The lawsuit accuses Anthropic of using UMG's songs to make new music. Anthropic's website states that its AI models can create original music in different genres and styles, depending on user choices. Users can also send their own audio files and have the AI models change them or add new features. The lawsuit claims that Anthropic's service relies on UMG's songs, which are legally protected.
Writers Are Also Trying to Protect Their Work from AI
A group of writers, including major figures like Michael Chabon and David Henry Hwang, have filed a lawsuit against OpenAI. They claim that the company unlawfully accesses their copyrighted works to train its AI model, ChatGPT. Chabon and the group have also brought a similar lawsuit against Meta Inc. for the same reasons.
Earlier in the year, Sarah Silverman, Christopher Golden, and Richard Kadrey accused both OpenAI and Meta of copyright infringement. They claim technology companies obtained their books from illegal sources, such as websites that offer free downloads of pirated books.
In July, a group of leading news publishers also considered suing AI companies over copyright infringement. The publishers allege that the AI firms are infringing on their intellectual property rights and undermining their business model by scraping, summarizing, or rewriting their articles and distributing them on various platforms, such as websites, apps, or social media.