Getty Images vs Stability AI: Key Legal Clash on AI and Copyright

Published Date : 17/06/2025 

The High Court trial of Getty Images v Stability AI is set to address the complex legal issues surrounding AI and copyright infringement, with far-reaching implications for the tech and media industries. 

The eagerly anticipated High Court trial in the case of Getty Images v Stability AI began on June 9, 2025. This case is expected to have significant implications for the use of copyrighted content by artificial intelligence technologies and will shape the future of intellectual property rights in the AI industry.

Both parties are well-known in their respective fields. Getty Images, the claimant, is a leading visual media company and supplier of stock images, editorial photography, video, and music, with a vast library of content. Stability AI, the defendant, is an artificial intelligence company known for image generation via its AI model, Stable Diffusion. Stable Diffusion produces images based on text or image prompts input by the user.

The first key issue before the High Court is copyright infringement. Getty Images alleges that Stability AI used millions of Getty's copyrighted content to train its AI model without permission. Getty claims that Stability used links in Getty's dataset to obtain Getty content from websites, storing and copying this content to be used in the training. Getty argues that copying without consent infringes Getty's exclusive rights to reproduce Getty content, under section 17 of the Copyright, Designs and Patents Act 1988 (CDPA). Stability AI accepts that some of Getty's content was used during the training but denies that this amounts to copyright infringement, contending that the training took place outside the UK and thus no act conflicting with Getty's intellectual property rights occurred within the UK.

Getty further alleges that some of the works produced by Stability's AI model are markedly similar to the copyrighted content used to train it, with some even possessing Getty's watermarks. Getty contends that this infringes on Getty's exclusive rights to reproduce the content and communicate it to the public, as well as the exclusive right to grant authority to reproduce or communicate the content. Getty argues that these infringements occur under sections 16(2) and 20 of the CDPA. Getty also argues that Stability's AI model is an infringing copy of Getty's content, of which Stability is responsible for importing into the UK, in breach of sections 22 and 23 of the CDPA.

Stability AI's position is that the content produced by its AI model only bears coincidental resemblance to Getty content and that the content produced by the AI model is not the responsibility of Stability but is instead down to the users of the model. Stability further contends that reasonable steps have been taken to prevent any infringements and that no infringements have been authorized.

The second key issue for determination is trademark infringement and passing off. Getty Images claims to have UK registered trademarks which appear as a watermark on Getty's content, and that these trademarks are linked to Getty's reputation due to the volume of Getty's users and marketing campaigns within the UK. Getty claims that Stability AI used Getty's trademarks to train its AI model without authority and that the content produced by the AI model bears Getty's trademarks. They argue that this amounts to deceptive trade practices and taking unfair advantage of Getty's trademark rights.

Stability AI denies this claim and contends that any watermarks produced by the AI model are not used by Stability nor are they an indicator of trademark use.

The third issue is database right infringement, as Getty asserts that Stability AI has extracted content from Getty's database without permission and used this content to train its AI model. To support this claim, Getty has detailed the investment made into the database, including obtaining and verifying the contents. Stability AI argues that Getty has not sufficiently demonstrated its database rights to claim that these rights have been infringed.

The result of this case is highly anticipated by a number of industries, especially content creators and AI software providers. Across the pond, Stability AI is also being sued in California for copyright infringement in an action brought by a number of artists. This case (and other similar actions) will likely provide clarity on the boundaries of using copyrighted material in AI model training and the management and protection of intellectual property. It will likely also shape the UK Government's proposals and policies going forward concerning the use of AI and third-party content.

If Getty Images successfully proves copyright infringement, there would be significant ramifications for the AI industry and all industries relying on such software. It is unclear whether AI developers can (in the absence of a full shutdown of the model) exclude or limit the content the model utilizes to produce its output. The High Court's decision will likely be sought to be appealed all the way to the UK Supreme Court and result in legislative intervention thereafter. 

Frequently Asked Questions (FAQS): 

Q: What is the main issue in the Getty Images v Stability AI case?

A: The main issue is copyright infringement, where Getty Images alleges that Stability AI used millions of Getty's copyrighted content to train its AI model without permission.

Q: What is Stability AI's defense against the copyright infringement claim?

A: Stability AI argues that the training took place outside the UK, and thus no act conflicting with Getty's intellectual property rights occurred within the UK.

Q: What are the potential implications of this case for the AI industry?

A: The case could have significant ramifications for the AI industry, especially regarding the use of copyrighted material in AI model training and the management and protection of intellectual property.

Q: What other legal issues are being addressed in this case?

A: The case also addresses trademark infringement and passing off, as well as database right infringement.

Q: What is the expected outcome of the case?

A: The outcome of the case is highly anticipated, and it is expected to provide clarity on the boundaries of using copyrighted material in AI model training and may lead to legislative intervention. 

More Related Topics :