AI art tools Stable Diffusion and Midjourney targeted in copyright lawsuit

A trio of artists has filed a lawsuit against Stability AI and Midjourney, makers of AI art generators Stable Diffusion and Midjourney, and artist portfolio platform DeviantArt, which recently created its own AI art generator DreamUp.

The artists – Sarah Andersen, Kelly McKernan and Karla Ortiz – allege that these organizations have violated the rights of “millions of artists” by training their AI tools on five billion images scraped from the internet “without the permission of the original artists”.

The lawsuit was brought by attorney and typographer Matthew Butterick along with the Joseph Saveri Law Firm, which specializes in antitrust and class action cases. Butterick and Saveri are currently suing Microsoft, GitHub and OpenAI in a similar case involving the AI ​​programming model CoPilot, which is trained on lines of code collected from the internet.

In a blog post announcing the suit, Butterick describes the case as “another step towards making AI fair and ethical for everyone”. He says the ability of AI art tools like Stable Diffusion to “flood the market with an essentially unlimited number of infringing images will do permanent damage to the art and artist marketplace.”

Since AI art tools have exploded in popularity over the past year, the art community has responded strongly. While some say these tools can be useful, just like previous generations of software like Photoshop and Illustrators, many more object to using their work to train these money-making systems. Generative AI art models are trained on billions of images collected from the internet, usually without the creators’ knowledge or consent. AI art generators can then be used to create works of art that replicates the style of specific artists.

Whether or not these systems infringe copyright law is a complicated question that experts say will have to be resolved in the courts. The creators of AI art tools generally claim that the training of this software on copyrighted data (at least in the US) is covered by fair use doctrine. But fair use cases have yet to be adjudicated and there are plenty of complicating factors when it comes to AI art generators. These include the location of organizations behind these tools (since the EU and US have subtly different legal fees for data scraping) and the purpose of these settings (e.g. Stable Diffusion is trained on the LAION datasetfounded by a non-profit research organization based in Germany, and non-profit organizations may be treated more favorably than mainstream companies in fair use cases).

The lawsuit brought by Butterick and the law firm of Joseph Saveri has also criticized for contain technical inaccuracies. For example, the lawsuit alleges that AI art models “store compressed copies of [copyright-protected] train images” and then “recombine” them; function as a “collage tool of the 21st century[s].” However, AI art models don’t store images at all, but rather mathematical representations of patterns gleaned from these images. The software also does not put pieces of images together in the form of a collage, but creates images from scratch based on these mathematical representations.

The edge has reached out to Matthew Butterick, Stability AI, Midjourney and DeviantArt for comment. We’ll update the story if we hear anything.