Stability AI largely wins UK court battle against Getty Images over copyright and trademark

LONDON (AP) — Artificial intelligence company Stability AI mostly prevailed against Getty Images Tuesday in a British court battle over intellectual property.

Seattle-based Getty had accused Stability AI of infringing its copyright and trademark by scraping 12 million images from its website, without permission, to train its popular image generator, Stable Diffusion.

The closely followed case at Britain’s High Court was among the first in a wave of lawsuits involving generative AI as movie studios, authors and artists challenged tech companies’ use of their works to train AI chatbots.

Tech companies have long argued that “fair use” or “fair dealing” legal doctrines in the United States and United Kingdom allow them to train their AI systems on large troves of writings or images. Tuesday's ruling provides some clarity but still leaves big unanswered questions over copyright and AI, experts said.

According to the judge’s written ruling, Getty narrowly won its argument that Stability had infringed its trademark, but lost the rest of its case.

Both sides claimed victory.

“This is a significant win for intellectual property owners,” Getty Images said in a statement.

Shares of Getty dipped 3% before the opening bell in the U.S.

Stability said it was pleased with the ruling.

“This final ruling ultimately resolves the copyright concerns that were the core issue,” Stability's General Counsel Christian Dowell said.

Getty had accused Stability of both primary and secondary copyright infringement.

Legal experts said the first one involves the act of reproducing something without permission — similar to a dodgy factory churning out counterfeit Chanel handbags or pirated CDs — while the second involves importing those copies from another country.

In this case, Getty said Stability's use of its image library to train and develop Stable Diffusion’s AI model amounted to breach of primary copyright. Stability responded that the case doesn’t belong in the United Kingdom because the AI model’s training technically happened elsewhere, on computers run by U.S. tech giant Amazon.

During the three-week trial in June, Getty dropped its primary copyright allegations, in a sign that it didn't think they would succeed. But it still pursued the secondary infringement claims. Even if Stability’s AI training happened outside the U.K., Getty said offering the Stable Diffusion service to British users amounted to importing unlawful copies of its images into the country.

Justice Joanna Smith rejected Getty’s claims, ruling that Stable Diffusion’s AI didn’t infringe copyright because it doesn’t “store or reproduce any Copyright Works (and has never done so).”

Getty also sued for trademark infringement because its watermark appeared on some of the images generated by Stability's chatbot.

The judge sided with Getty but added that part of the case only partially succeeded, and that her findings are "both historic and extremely limited in scope."

“While I have found instances of trademark infringement, I have been unable to determine that these were widespread," she said.

Experts said Getty’s decision to drop part of its copyright case means AI training is still in legal limbo.

“The decision leaves the U.K. without a meaningful verdict on the lawfulness of an AI model’s process of learning from copyright materials,” said Iain Connor, an intellectual property partner at law firm Michelmores.

Smith said there was "very real societal importance" in deciding how to strike a balance between the creative and tech industries. But she added that the court can only rule on the "diminished" case that remained and couldn't consider “issues that have been abandoned.”

A Getty spokeswoman declined to say whether there would be an appeal.

Getty is also pursuing a copyright infringement lawsuit in the United States against Stability. It originally sued in 2023 but refiled the case in a San Francisco federal court in August.

The Getty lawsuits are among a slew of cases that highlight how the generative AI boom is fueling a clash between tech companies and creative industries.

Anthropic agreed to pay $1.5 billion to settle a class-action lawsuit by book authors who say the company took pirated copies of their works to train its Claude chatbot.

Separately, a federal judge dismissed a lawsuit from a group of 13 authors who made similar accusations against Facebook owner Meta Platforms in training its AI system Llama.

Warner Bros. has sued Midjourney for copyright infringement, alleging that its image generator enables subscribers to create AI-generated images and videos of copyrighted characters like Superman and Bugs Bunny.