The image repository claims Stability AI ‘unlawfully copied’ millions of copyright-protected images to train Stable Diffusion, without seeking a licence.
Getty Images is suing Stability AI for allegedly stealing copyrighted content for its AI art creator, Stable Diffusion.
The stock image supplier claims Stability AI “unlawfully copied and processed” millions of copyright-protected images for its own commercial benefit and “to the detriment of content creators”.
Getty Images said it has commenced legal proceedings in London’s High Court of Justice against Stability AI.
Stable Diffusion is a text-to-image generator that can create new images based on a sentence a user inputs, combining different concepts, attributes and styles. These AI models like Midjourney and DALL-E are trained using a massive amount of existing images.
However, concerns have been raised that some of the content used to train these models could be copyright protected and taken without the consent of artists.
Last year, an analysis of 12m images used to train Stable Diffusion found that around 47pc were sourced from only 100 domains, with the largest number of images (around 8.5pc) coming from Pinterest. This analysis also suggested that some of these training images could be copyright protected.
Last September, Getty Images issued a ban on the upload and sale of AI-generated images on its platform. The company said this was due to “open questions” surrounding the copyright of AI-generated images, along with uncertainty surrounding the data these AI models are trained on.
In a statement on its legal proceedings, Getty Images said it has provided licences to “leading technology innovators” for the purpose of training AI systems in a way that “respects personal and intellectual property rights”.
“Stability AI did not seek any such licence from Getty Images and instead, we believe, chose to ignore viable licensing options and long‑standing legal protections in pursuit of their standalone commercial interests,” Getty Images said.
AI-generated images became prevalent last year as various text-to-image models surged in popularity. These models offer new possibilities for users – including using their generations for commercial purposes.
But legal questions have been raised, such as who truly owns the images and if they might infringe on existing copyrighted works.
Last September, JumpStory co-founder Jonathan Løw told SiliconRepublic.com that there is a growing “legal minefield” around using AI-generated images for commercial purposes. He also said that the legal risk may fall on the end user if their commercially used image enters a copyright dispute.
This week, a group of artists launched a lawsuit against Stability AI and Midjourney for allegedly taking the work of “millions of artists” without their consent, The Verge reports.
10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.