AI legal battles: U.K. trial and new U.S. lawsuit unfold
Getty Images fights Stability AI over copyright infringement in the U.K. courts, while Reddit sues Anthropic for contract breach over unauthorized data scraping for AI training.
A set of new lawsuits is illuminating the main friction between generative AI vendors and those accusing them of using their content without permission.
Opening arguments began Monday at the British High Court in the lawsuit filed by Stability AI. Getty Images accuses the generative AI (GenAI) imaging vendor of infringing on its copyrighted materials by using Getty's photography collection to train its Stable Diffusion image-generating model.
The London trial started five days after social media company Reddit sued GenAI vendor Anthropic in the U.S. in a similar case that is awaiting trial.
Reddit vs Anthropic
On June 4, Reddit sued the vendor for breach of contract. The social media company said the AI vendor is training its models on the personal data of Reddit users without permission. Reddit said that while other AI vendors such as OpenAI and Google use Reddit's content, both have agreed to Reddit's licensing terms.
The Reddit vs. Anthropic lawsuit differs from the Getty vs. Stability lawsuit. The Getty action concerns copyright infringement, while the Reddit suit alleges breach of contract.
"This is effectively a contract law fight," said Michael G. Bennett, associate vice chancellor for data science and AI strategy at the University of Illinois Chicago, referring to the Reddit suit. He added that Reddit is saying that anyone who uses its website agrees to its terms of use, which prohibits the way it believes Anthropic uses the website.
It's highly invested in both signaling to society at large and these huge AI companies in particular that the better route for them is to come to the negotiating table and quickly make an agreement.
Michael G. BennettAssociate vice chancellor for Data Science and AI strategy, University of Illinois Chicago
"To the extent that the court relies heavily or even exclusively on that agreement language that we all agree to upon accessing and making use of Reddit content, I think that that distinction that Anthropic wants to make may be of minimal value," Bennett said.
Moreover, it seems clear that Reddit wants to continue negotiations with companies that want to use its models to train AI content. Therefore, Bennett continued, a company like Anthropic that is allegedly using its content without paying blows up the social media business model.
"It's highly invested in both signaling to society at large and these huge AI companies in particular that the better route for them is to come to the negotiating table and quickly make an agreement and get access to all of that Reddit content that these companies think is valuable and worth integrating into their training processes," Bennett said.
Getty Images, Stability AI and two countries
While the lawsuit between Getty Images and Stability AI concerns copyright infringement and not contract law, Getty also says that gaining compensation from AI vendors that seek to use its content to train their models is important.
Getty's trial lawyers argued in the U.K. court that the case isn't about a battle between AI technology and the creative industry; rather, it's about the problem created when AI companies use creative works to train their models without payment.
While Stability is arguing fair use, the outcome of the case in the U.K. will likely differ from its outcome in the U.S., where the fair use argument is more flexible. Fair use is the part of copyright law that allows limited use of copyrighted material without permission from the creators of that material.
"You could very well see the U.K. court finding that there is a copyright infringement, and it doesn't fall in any of these exceptions outlined under the U.K. laws, and then the U.S. court reaching a different decision," said Vincent Allen, partner at intellectual property law firm Carstens, Allen & Gourley.
Moreover, the U.K. trial will be decided by a judge, while the U.S. trial could be decided by a jury.
It's unclear if the U.K. trial will affect the U.S. trial, Allen said.
"There could be some factual determinations that the judge makes in that case that might be considered persuasive in the U.S., but I think ultimately it's going to be an independent decision process," he said.
He added that no matter each country's court decision, the losing party will likely still have the chance for protracted appeals.
"There's going to be a long road ahead when it comes to the appeal of these fair use decisions, and ultimately, the Supreme Court will probably weigh in on what is considered fair use in the AI training contest," Allen continued. "It may be several years before we ever have any finality on this issue unless Congress were to act sooner and enact some related legislation."
However, Bennett said the current industry would greatly benefit if these issues about fair use were decided sooner rather than later.
Esther Shittu is an Informa TechTarget news writer and podcast host covering artificial intelligence software and systems.
