Getty Images is suing Stability AI for allegedly stealing its photos to train a machine. And it’s not a small spat. This could be the case that rewrites how copyright law handles artificial intelligence.
Kicking off yesterday, 9 June 2025, in London’s High Court, Getty’s lawsuit accuses Stability AI of grabbing over 12 million copyrighted images without permission. The images were allegedly used to train Stable Diffusion, the text-to-image generator that’s sparked a thousand headlines and almost as many lawsuits.
It’s not Getty’s first rodeo. The company is already pursuing a similar case against Stability AI in the U.S. This one, though, could be the watershed moment for AI and copyright law.
A Line in the Sand for Creators
Getty’s message is clear: they see this as theft. Plain and simple. Craig Peters, Getty’s CEO, has condemned the practice as a violation of copyright and a blow to the creative industry. He’s not mincing words.
Meanwhile, Stability AI insists it’s playing by the rules. It argues that training AI on public data is fair use. That it reflects collective human knowledge. That it doesn’t steal; it learns.
But Getty’s not buying it. And neither are a growing number of rights holders watching AI chew through their content with little to no compensation.
Copyright in the Age of AI
This is more than a single case. It’s part of a global reckoning.
In the U.S., judges have already let copyright infringement claims proceed against Stability AI in a lawsuit brought by artists. The courts are starting to take a hard look at whether these models are just remixing, or outright copying, protected works.
And the fair use defense? It’s looking shaky. Especially when the AI outputs are so close to the real thing.
A Legal Maze with No Clear Map
What makes this even trickier is the patchwork of global laws. In the U.S., AI-generated work can’t be copyrighted unless there’s a human author behind it. In other places, AI-assisted content might still get protection.
For companies like Stability AI, this means navigating a minefield. One wrong step in one country could trigger lawsuits around the globe. Compliance becomes a nightmare.
Getty’s Bigger Play
This lawsuit isn’t just about one model or one batch of images. It’s about control. About setting a precedent. Getty has been both the hunter and the hunted in copyright battles before. Now, it’s going all in.
As AI models grow more powerful, traditional content owners are shifting gears. They’re no longer sitting on the sidelines. They’re building legal firewalls. And they’re ready to fight.
Dr. Ilia Kolochenko, CEO at ImmuniWeb and Fellow at the British Computer Society (BCS), weighs in:
“While Getty’s case will certainly bring some clarity whether copyright law prohibits or restricts unwarranted use of creative content (such as texts, images, videos) for AI training purposes, it will unlikely cause a tectonic shift in the AI industry. Fierce legal battles currently take place in courts on both sides of the Atlantic disputing the alleged copyright infringements by numerous AI vendors. Ironically, at the end of the day, these lawsuits are somewhat futile: virtually all creative content providers have already incorporated paywalls and advanced anti-bot protection mechanisms on their websites, as well as added specific contractual provisions to their terms of service that expressly prohibit any use of their data for LLM training purposes.
He says in case of a violation of the terms of service, content providers will have a straightforward claim for breach of contract, possibly accompanied by liquidated damages per violation, making such claims extremely lucrative for the plaintiffs. Furthermore, in some jurisdictions, a deliberate bypass of anti-bot protection and massive data scraping may constitute a criminal offense.
“And all this has virtually nothing to do with copyright law. Ultimately, AI vendors, that now vigorously argue in courts that exploitation of third-party content for LLM training purposes constitutes a fair use exception under copyright law, will likely face even greater liability under the avalanche of breach of contract claims.”
A Tough Reality
According to him, regardless of the outcome of the Getty’s case, most AI vendors will soon face a tough reality: paying a fair price for high-quality training data, while staying profitable. In view of the formidable competition emanating from China, many Western AI companies may simply quit the business as economically unviable.
Kolochenko’s last point cuts deep. The economics of AI may hit a wall, one paved with licensing fees, legal battles, and territorial copyright rules. And if vendors can’t afford to train their models ethically? They may fold.
Information Security Buzz News Editor
Kirsten Doyle has been in the technology journalism and editing space for nearly 24 years, during which time she has developed a great love for all aspects of technology, as well as words themselves. Her experience spans B2B tech, with a lot of focus on cybersecurity, cloud, enterprise, digital transformation, and data centre. Her specialties are in news, thought leadership, features, white papers, and PR writing, and she is an experienced editor for both print and online publications.
The opinions expressed in this post belong to the individual contributors and do not necessarily reflect the views of Information Security Buzz.