Background
Getty Images (US) Inc, Getty Images International UC, Getty Images (UK) Limited, Getty Images Devco UK Limited, iStockphoto LP and Thomas M Barwick Inc brought claims against Stability AI Ltd over use of Getty content in training Stable Diffusion and over allegedly infringing outputs. This judgment is a procedural ruling on a case-management issue: the court disallowed a proposed representative action by the sixth claimant, so the case continues toward trial on the remaining issues.
AI interaction
‘Standing back, and trying to see the wood for the trees, it is plain that the question of which Copyright Works have been used for training Stable Diffusion is not a question that can currently be determined.’ The court records the claim that Stable Diffusion was trained on scraped Getty images and that its outputs may reproduce protected works or bear Getty trade marks, and highlights the practical challenge of identifying training data at scale; that challenge was central to rejecting the representative action proposal.
Update Note (November 2025 – [2025] EWHC 2863 (Ch))
In a subsequent judgment, the court ordered targeted disclosure concerning the Stable Diffusion model, including “training data”, “datasets”, and the processes used to generate model weights, finding that this material is relevant to determining the pleaded issues. This moves the case beyond procedural management toward substantive assessment of AI-training conduct and strengthens expectations of transparency around model training in UK litigation.
It is significant because it sets a UK precedent requiring transparency about model training and datasets in generative-AI copyright cases.