According to Financial Times News, Getty Images lost a central part of its UK copyright lawsuit against Stability AI in a High Court ruling on Tuesday. Mrs Justice Joanna Smith ruled that Stable Diffusion isn’t an “infringing copy” since it doesn’t store or reproduce copyright works. Getty dropped its original copyright claim after finding no evidence that Stable Diffusion’s training happened in the UK, instead pursuing narrower secondary infringement claims. The decision comes after recent US cases where Meta and Anthropic won similar copyright battles, with courts finding their use of copyrighted materials for AI training qualified as “fair use.” Getty’s shares fell about 5 percent following the ruling, though the company did secure a win on unauthorized use of its trademarks when Stable Diffusion generated images with Getty watermarks.
A “massive damp squib” for creators
Here’s the thing: this was supposed to be the big one. Creative industries have been watching this case closely, hoping for clarity on whether AI companies can just scoop up their work without permission or payment. And what did they get? Lawyers are calling it a “massive damp squib” – basically a fancy way of saying it fizzled out.
The ruling doesn’t actually answer the billion-dollar question: is training AI on copyrighted materials legal in the UK? Getty’s decision to drop key parts of their case meant the court never had to rule on that fundamental issue. So we’re left with this weird situation where the AI companies get a win, but nobody really knows what the rules are.
What actually was decided
Look, the court did make some concrete findings. Stability AI can’t generate images with Getty’s watermarks – that’s trademark infringement. But the core copyright question? Basically untouched. The judge focused on the technical fact that Stable Diffusion doesn’t literally store copies of Getty’s images, which makes sense from a pure copyright perspective.
But here’s what’s interesting: Getty says they’ll use these findings in their US case. So this UK ruling might become ammunition in a different legal battle. It’s like losing a skirmish but planning to use the intelligence in the main war.
The bigger picture for AI and creativity
This feels like part of a pattern, doesn’t it? US courts siding with AI companies, UK courts delivering limited rulings that don’t settle the big questions. Creative industries are getting increasingly frustrated while AI companies are getting more confident.
Robert Guthrie from Osborne Clarke called this “a big win for Stability AI and AI developers generally.” He’s probably right in the short term. But Catriona MacLeod Stevenson from the Publishers Association says this isn’t the end of the road. The fundamental tension between AI training and creator rights isn’t going away anytime soon.
So where does this leave us? Basically back where we started, but with more legal bills and less clarity. The creative industries want compensation when their work trains systems that might eventually replace them. AI companies want to train their models without drowning in licensing costs. And the courts? They’re still figuring out how century-old copyright laws apply to technology that didn’t exist five years ago.
