By Gary Symons
TLL Editor in Chief
In a world first, Getty Images is suing AI art generator Stability AI for copyright infringement, alleging it processed millions of its images.
Stability AI is the company that created the popular art generation tool Stability Diffusion, which has already created millions of images for its users. However, like other similar programs, the artificial intelligence is ‘trained’ using images from other sources. Getty alleges many of those images were obtained from its database.
“This week Getty Images commenced legal proceedings in the High Court of Justice in London against Stability AI, claiming Stability AI infringed intellectual property rights including copyright in content owned or represented by Getty Images,” the company said in a brief statement. “It is Getty Images’ position that Stability AI unlawfully copied and processed millions of images protected by copyright and the associated metadata owned or represented by Getty Images, absent a license, to benefit Stability AI’s commercial interests and to the detriment of the content creators.”
Getty’s issue with Stability AI is not that the young artificial intelligence company is using its images, but rather that it failed to acquire the necessary license.
Getty has in fact worked with other AI image generator firms. For example, in October, Getty announced it had formed a partnership with BRIA, developer of proprietary AI visual content tools. The goal is to allow creatives to transform images to their specific needs using intuitive AI tools on Getty Images’ platform.
But in the case of Stability AI, Getty says the company failed to acquire or even seek a license.
“Getty Images believes artificial intelligence has the potential to stimulate creative endeavors,” the company said. “Accordingly, Getty Images provided licenses to leading technology innovators for purposes related to training artificial intelligence systems in a manner that respects personal and intellectual property rights.
“Stability AI did not seek any such license from Getty Images and instead, we believe, chose to ignore viable licensing options and long‑standing legal protections in pursuit of their stand‑alone commercial interests.”
Getty Images announced a lawsuit against Stability AI, the company behind popular AI art tool Stable Diffusion, alleging the tech company committed copyright infringement.
The stock image giant accused Stability AI of copying and processing millions of its images without obtaining the proper licensing, according to a press release issued Tuesday. London-based Stability AI announced it had raised $101 million in funding for open-source AI tech in October and released version 2.1 of its Stable Diffusion tool in December.
Opinion: Artists Are Freaking Out Over AI Apps, And So Should You
“Getty Images believes artificial intelligence has the potential to stimulate creative endeavors. Accordingly, Getty Images provided licenses to leading technology innovators for purposes related to training artificial intelligence systems in a manner that respects personal and intellectual property rights,” Getty wrote in the statement. “Stability AI did not seek any such license from Getty Images and instead, we believe, chose to ignore viable licensing options and long standing legal protections in pursuit of their stand-alone commercial interests.”
Stability AI has not yet issued a full statement about the legal action, but has released a number of ‘no comment’ statements to media outlets. The gist of their comments so far is that they have not yet been served legal documents.
“Please know that we take these matters seriously. It is unusual that we have been informed about this intended legal action via the press,” a Stability AI spokesperson told CNN. “We are still awaiting the service of any documents. Should we receive them, we will comment appropriately.”
However, in a previous story on CNN in October 2022, prior to the lawsuit, Stability AI founder and CEO Emad Mostaque said via and email that art is a tiny fraction of the LAION training data behind Stable Diffusion.
“Art makes up much less than 0.1% of the dataset and is only created when deliberately called by the user,” Mostaque said.
This is the first known lawsuit by an image licensor like Getty against artificial intelligence-powered image generators, but there have been other legal clashes in the uneasy relationship between human content creators and the AI constructs that are trained using their works.
This lawsuit is essentially an escalation in the developing legal battles between AI firms and human content creators for credit and payment related to the use of images, and the result will determine the landscape of the creative industries not just for imagery, but for writing, music and video as well.
The issue is that AI art tools rely on human-created artwork to both train and to essentially get the style and the idea for a new image. The AI firms claim this practice is covered by laws in various companies, such as the American legal concept of Fair Use.
But many content creators do not agree and argue it constitutes copyright violation. In early January, for example, three artists filed a lawsuit against Stable Diffusion and Midjourney, as well as the artist portfolio platform DeviantArt, which recently created its own AI art generator, called DreamUp.
Plaintiffs Sarah Andersen, Kelly McKernan, and Karla Ortiz allege these organizations infringed the rights of “millions of artists” by training their AI tools on five billion images scraped from the web “without the consent of the original artists.”
While Stability AI has said the use of images is not a major part of the Stable Diffusion system, it is known that the AI bot has vacuumed up millions or perhaps billions of images.
Two data experts, Andy Baio and Simon Willison, recently published an article on a project designed to show where some of these images came from.
“We grabbed the data for over 12 million images used to train Stable Diffusion, and used (Willison’s) Datasette project to make a data browser for you to explore and search it yourself,” wrote Baio. “Note that this is only a small subset of the total training data: about 2% of the 600 million images used to train the most recent three checkpoints, and only 0.5% of the 2.3 billion images that it was first trained on.”
The pair indexed the 12 million images by domain, and found that nearly half of the images, about 47%, were sourced from only 100 domains, with the largest number of images coming from Pinterest. Over a million images, or 8.5% of the total dataset, are scraped from Pinterest’s pinimg.com Content Distribution Network, or CDN.
They also found that user generated content sites were a major source in general, including WordPress blogs at 6.8%; other photo and art sites were significant content sources including:
- 232k images from Smugmug
- 146k from Blogspot
- 121k images from Flickr
- 67k images from DeviantArt
- 74k from Wikimedia
- 48k from 500px
- 28k from Tumblr
Others came from shopping sites, the biggest being Fine Art Ameria with 698,000 images, or about 5.8% of the 12 million images, as well as large amounts from Shopify, Wix, Squarespace, Redbubble, and Etsy.
Significantly for the Getty Images lawsuit, the researchers said “… a large number came from stock image sites. 123RF was the biggest with 497k, 171k images came from Adobe Stock’s CDN at ftcdn.net, 117k from PhotoShelter, 35k images from Dreamstime, 23k from iStockPhoto, 22k from Depositphotos, 22k from Unsplash, 15k from Getty Images, 10k from VectorStock, and 10k from Shutterstock, among many others.”
While a relatively small number came from Getty, this project only included 12 million out of more than 2 billion images in the Stability Diffusion database.
Also, while Getty and some content creators are launching lawsuits over copyright, it appears that most image licensors are willing to work with AI art generators.
For example, Getty competitor Shutterstock announced plans in October to expand its partnership with OpenAI, the company behind DALL-E and viral AI chat bot ChatGPT, and enhance AI-generated content while launching a fund to compensate artists for their contributions.