See How Visible Your Brand is in AI Search Get Free Report

Getty Images vs Stability AI: Getty Images Loses to Stability AI in Landmark UK Lawsuit Over AI Copyright

  • November 4, 2025
    Updated
getty-images-vs-stability-ai-getty-images-loses-to-stability-ai-in-landmark-uk-lawsuit-over-ai-copyright

A London judge handed Stability AI a mostly favorable result while giving Getty Images a narrow trademark win, leaving the biggest copyright question about AI unresolved in the UK.

📌 Key Takeaways

  • Judge upheld a limited trademark claim over Getty-style watermarks in some outputs.
  • Court said Stable Diffusion is not an “infringing copy” because it does not store works.
  • Getty dropped core training claims mid-trial, so UK law on training remains open.
  • Both sides claimed wins, but the balance of findings favors Stability AI.
  • Ruling sparked calls for clearer transparency and copyright guidance from government.


What The Court Decided — And What It Didn’t

The judge found some trademark infringement when outputs contained Getty-like watermarks, yet called those findings historic and extremely limited in scope. The rest of Getty’s claims largely fell away after the company withdrew key training allegations.

Crucially, the court said an AI model that does not store or reproduce copyrighted works is not an “infringing copy.” That left the high-stakes question, whether training on copyrighted works is infringement in the UK, undecided.

“An AI model such as Stable Diffusion which does not store or reproduce any copyright works is not an ‘infringing copy.’” — Mrs Justice Joanna Smith

A practical implication follows. Future cases will need clean evidence about where training happened and how models represent data, or courts will continue to sidestep the core training issue.


Why Each Side Says It Won

Getty highlighted that the court recognized trademark issues when watermarks appear in outputs and argued responsibility sits with the model provider, not end users. The company framed that as a win for rights holders and a lever for U.S. litigation.

Stability AI emphasized the dismissal of secondary copyright claims and the court’s statement that Stable Diffusion is not an infringing copy. For Stability, that resolves the “core” copyright concern raised.

“Responsibility for the presence of such trademarks lies with the model provider. This is a significant win for intellectual property owners.” — Getty Images

That split narrative explains the headlines, with Getty pointing to brand protection and Stability spotlighting the copyright backbone.


What It Means For Creators, Models, And Policymakers

For creators, the ruling is a reminder that watermarks and brand signals still matter in court, but protection against training remains in legal limbo. Without transparency rules, rights holders face costly, uncertain fights.

For model builders, the decision underscores the value of provenance records and explicit controls that suppress watermarks and brand artifacts. It also raises the bar for demonstrating where training occurred and how data was handled.

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

Lawyers now expect pressure on the UK to clarify whether text and data mining for training is allowed by default or only with opt-out or consent, and to define the disclosures models must keep.


Market And Industry Reaction

The mixed outcome dented Getty’s stock in premarket trading, reflecting investor uncertainty about legal leverage. At the same time, analysts noted the ruling could embolden AI firms, at least while the training question stays open.

Expect more suits to test different facts, plus policy moves on dataset transparency, watermark controls, and record-keeping. Until lawmakers act, case outcomes will hinge on evidence about pipelines and model internals.


Conclusion

The court drew a narrow line. Trademark issues around watermarks can stick, but modeling that does not store works is not an infringing copy. The headline fight over training on copyrighted data remains unanswered.

Without clearer rules, creators and AI companies will keep meeting in court. The next decisive step likely comes from Parliament or a case with airtight evidence on where and how models are trained.


📈 Trending News

  • Channel 4 Debuts UK’s First AI Presenter That Fooled Viewers
  • Edinburgh’s EPCC to Run UK AI Factory Antenna with €10m Funding
  • eBay Launches ‘AI Activate’ to Bring ChatGPT Enterprise to 10,000 Sellers
  • £500M Recovered: How AI Helped the UK Government Catch Fraudsters
  • Nory Raises £27M: Can These AI Assistants Run Restaurants on Their Own?

For more news and insights, visit AI News on our website.

Was this article helpful?
YesNo
Generic placeholder image
Articles written 25

Khurram Hanif

Reporter, AI News

Khurram Hanif, AI Reporter at AllAboutAI.com, covers model launches, safety research, regulation, and the real-world impact of AI with fast, accurate, and sourced reporting.

He’s known for turning dense papers and public filings into plain-English explainers, quick on-the-day updates, and practical takeaways. His work includes live coverage of major announcements and concise weekly briefings that track what actually matters.

Outside of work, Khurram squads up in Call of Duty and spends downtime tinkering with PCs, testing apps, and hunting for thoughtful tech gear.

Personal Quote

“Chase the facts, cut the noise, explain what counts.”

Highlights

  • Covers model releases, safety notes, and policy moves
  • Turns research papers into clear, actionable explainers
  • Publishes a weekly AI briefing for busy readers

Related Articles

Leave a Reply