Getty's image-scraping sueball against Stability AI will go to trial in the UK
Computer-generated image maker fails to quash case on its claimed home turf
Getty's lawsuit against Stability AI, claiming the startup "unlawfully copied and processed millions of images protected by copyright" from its photo archive, will go to trial in the UK.
In January, Getty sued Stability and accused it of infringing upon the photo library's intellectual property rights and copyright protections by scraping its images without permission. Judges from the High Court of Justice in London believe that both sides should be given a chance to present evidence and arguments in court to resolve the issue.
These images were, it's claimed, used in datasets that helped train Stable Diffusion models. Getty believes Stability has avoided paying licenses to use its images, and instead unfairly profits off content creators. While Stability hasn't explicitly denied scraping the images, it believes the case is moot since the training process was completed using resources obtained outside of the UK.
The datasets used to train its model were seemingly compiled by LAION, a non-profit organization registered in Germany. Meanwhile, the hardware used to crunch the numbers during the process were from AWS servers in the US. CEO Emad Mostaque also claimed that none of the developers working to develop the models were based in the UK.
But the High Court judges aren't entirely convinced. "Having examined with care all the evidence before the court, I am not so satisfied," justice Joanna Smith ruled in court documents [PDF]. Smith reckoned that there were still some "unanswered questions" and "inconsistencies" in Stability's arguments, and that Getty's claims may hold up in court.
"All of this clearly supports the proposition that the Training and Development Claim has a real prospect of success and must be permitted to go to trial. The Location Issue is certainly not an issue on which I can say at present that the Claimants' claim is doomed to fail," she wrote.
- Author hopes to throw the book at OpenAI, Microsoft with copyright class action
- Microsoft dials back Bing after users manage to recreate Disney logo in fake AI-generated images
- AI copyright row deepens: Stability VP quits in protest over 'fair use' excuse
- To pay or not to pay for AI's creative 'borrowing' – that is the question
As for the copyright infringement claims, Smith argued it was a legal issue that ought to be straightened out at trial. "That point of law must be decided by the trial judge upon full and comprehensive argument from both sides and further to factual findings as to the nature of the acts said to give rise to the secondary infringement," she wrote.
A spokesperson representing Getty told The Register that the plaintiff is "pleased with the decision and look forward [to] the case being determined at trial." The stock image giant has filed a second lawsuit against Stability in the US.
Multiple lawsuits filed against Stability AI and other top developers of text-to-image tools or AI chatbots – like OpenAI, Google, Microsoft, and Anthropic – are still pending. Plaintiffs have accused developers of scraping training data – made up of books, art, code, song lyrics, and more – without consent, and violating copyright laws. ®