Key takeaways:
- After a US judge denied a class-action lawsuit against multiple generative AI corporations due to insufficient evidence, artists have suffered a blow in their copyright battle.
- The judge granted a class action member’s allegation of copyright infringement against Stability and gave the class thirty days to try to file an amended suit with more evidence.
After a US judge denied a class-action lawsuit against multiple generative AI corporations due to insufficient evidence, artists have suffered a blow in their copyright battle against the companies.
California District Court Judge William Orrick said in a ruling dated October 30 that the copyright infringement lawsuit against the AI company Stability AI, the art platform DeviantArt, and the generative AI image service Midjourney was “defective in numerous respects,” accepting the firms’ previous requests for their dismissal.
On the other hand, Judge Orrick granted a class action member’s allegation of copyright infringement against Stability and gave the class thirty days to try to file an amended suit with more evidence. As Orrick put it:
“Even Stability recognizes that determination of the truth of these allegations — whether copying in violation of the Copyright Act occurred in the context of training Stable Diffusion or occurs when Stable Diffusion is run — cannot be resolved at this juncture,”
The initial complaint was submitted in the middle of January, alleging that Stability’s artificial intelligence model, Stable Diffusion, was trained on billions of copyrighted photos, including those belonging to the artists, without authorization.
The lawsuit claimed that DeviantArt violated its own terms of service and may have copied millions of photos from the website without a license by incorporating Stable Diffusion on its platform.
Orrick stated that since it’s not “plausible” that the AI-generated graphics are taken from copyrighted photos, they probably don’t violate the artists’ copyright. The judge mentioned:
“Plaintiffs fail to allege specific plausible facts that DeviantArt played any affirmative role in the scraping and using of Anderson’s and other’s registered works to create the training images”
Until the class can demonstrate that the generated visuals are similar to the artists’ work, he continued, he is “not convinced.”
Some class members’ copyright claims were rejected because their photos weren’t registered with the Copyright Office, which is required to file a lawsuit alleging copyright infringement.
Similar lawsuits brought against AI companies, including the Author’s Guild class action against OpenAI, the Universal Music Group complaint against Anthropic, and the Getty Images litigation against Stability AI in the United States and the United Kingdom, are centered upon claims of copyright infringement.
