FILE PHOTO: Photographers are seen at the 62nd Cannes Film Festival, May 18, 2009. REUTERS/Regis Duvignau/File Photo

By Sam Tobin

LONDON (Reuters) -Getty Images largely lost its London lawsuit against artificial intelligence company Stability AI over its image generator on Tuesday, prompting Getty and some lawyers to call for stronger protections for copyright owners in Britain.

Seattle-based Getty, which produces editorial content and creative stock images and video, accused Stability AI of using its images to "train" its Stable Diffusion system, which generates images from text inputs.

The company had sued Stability AI for breach of copyright on the grounds Stable Diffusion was trained using Getty's images, and that images generated by Stable Diffusion reproduced its copyrighted images.

But Getty dropped that part of its case mid-trial, partly due to a lack of evidence about where Stable Diffusion was "trained", which two intellectual property lawyers said rendered Tuesday's ruling a "damp squib" for the law on AI.

JUDGE SAYS FINDINGS ARE LIMITED IN SCOPE

Getty pursued its claims of trademark infringement and secondary copyright infringement, alleging Stability AI imported into the United Kingdom an AI model which breached its copyright.

Judge Joanna Smith ruled Getty had succeeded "in part" on trademark infringement in relation to Getty watermarks generated by users of Stable Diffusion, but she said her findings were "both historic and extremely limited in scope".

She also dismissed Getty's secondary copyright infringement claim, on the grounds that "Stable Diffusion... does not store or reproduce any copyright works", which lawyers said exposed weaknesses in Britain's copyright protections.

"Today's finding means that copyright owners' exclusive right to reap what they have sown has been avoided on a technicality," Rebecca Newman, a lawyer at Addleshaw Goddard, said.

GETTY, STABILITY AI BOTH WELCOME RULING

The trial of Getty's case – one of several lawsuits brought in Britain, the U.S. and elsewhere over the use of copyright-protected material to train AI models – began in June with much fanfare about its potential wider impact on the law on AI.

Stability's lawyers argued Getty's lawsuit posed "an overt threat to ... the wider generative AI industry", a contention rejected by Getty which said it was simply upholding its intellectual property rights.

Getty Images said in a statement that the ruling "confirms that Stable Diffusion's inclusion of Getty Images' trademarks in AI-generated outputs infringed those trademarks", which Getty said was "a significant win for intellectual property owners".

The statement added that the ruling "established a powerful precedent that intangible articles, such as AI models, are subject to copyright infringement claims in the same way as tangible articles", which Getty said it would use in its parallel lawsuit against Stability AI in the U.S.

Stability AI welcomed the judgment, which its general counsel Christian Dowell said in a statement "ultimately resolves the copyright concerns that were the core issue".

Getty, whose shares were down 8.7% in premarket trading on the NYSE as of 1228 GMT, said the decision left it concerned that "even well-resourced companies such as Getty Images face significant challenges in protecting their creative works".

"We urge governments, including the UK, to establish stronger transparency rules which are essential to prevent costly legal battles and to allow creators to protect their rights," Getty said.

Gill Dennis, a lawyer at Pinsent Masons, said the ruling meant the government should now provide "clear, timely policy guidance" on the question of whether training AI on copyrighted material is an act of infringement.

Getty's decision to abandon large parts of its case means the issue of whether training AI models on copyrighted works constitutes infringement in the UK is unresolved, lawyers said.

Iain Connor, from the law firm Michelmores, said that "leaves the UK without a meaningful verdict on the lawfulness of an AI model's process of learning from copyright materials".

(Reporting by Sam Tobin; editing by Sarah Young, Alexandra Hudson)