Getty Images has largely lost its closely watched legal battle in London against Stability AI, the developer of the image generator Stable Diffusion, in a decision that highlights major gaps in British copyright law, according to Reuters.
The Seattle-based photo agency had accused Stability AI of unlawfully using its catalog of licensed images to train its artificial intelligence model, arguing that the resulting outputs reproduced protected content. However, Getty dropped much of its copyright claim mid-trial due to limited evidence showing where the model was trained, a move that legal experts say weakened its case.
Limited trademark victory, but no ruling on AI training legality
Judge Joanna Smith ruled that Getty had succeeded only in part—specifically on trademark infringement related to the appearance of Getty watermarks in images generated by users of Stable Diffusion. Still, the judge described the findings as “historic and extremely limited in scope.”
She dismissed Getty’s broader copyright infringement claims, concluding that “Stable Diffusion does not store or reproduce any copyright works.” Intellectual property lawyers said the decision exposed a significant loophole in how UK law treats the use of copyrighted materials for AI training.
“Today’s finding means that copyright owners’ exclusive right to reap what they have sown has been avoided on a technicality,” said Rebecca Newman, a lawyer at Addleshaw Goddard, after the ruling.
Both companies claim partial victories
While Getty called the decision “a significant win” for establishing that intangible digital systems can still fall under trademark rules, the company expressed concern that protecting creative content from AI misuse remains difficult even for major firms. Its shares fell nearly 9% in premarket trading on the New York Stock Exchange following the news.
“We urge governments, including the UK, to establish stronger transparency rules to help creators protect their rights,” Getty said in a statement.
Stability AI, for its part, welcomed the outcome. The company’s general counsel, Christian Dowell, said the judgment “resolves the copyright concerns that were the core issue” and provides clarity for the generative AI industry.
Legal uncertainty persists for AI training
The case, which began in June, was one of several global lawsuits testing whether training artificial intelligence models on copyrighted materials constitutes infringement. With Getty withdrawing the central portion of its claim, the question remains unresolved in the UK.
“This leaves the UK without a meaningful verdict on the lawfulness of an AI model’s process of learning from copyright materials,” said Iain Connor, a partner at Michelmores.
Lawyers and policy experts are now urging the British government to issue clear guidelines on AI data training and copyright, warning that uncertainty could trigger more costly disputes in the future.