Getty Images has begun legal proceedings against the creator of the Stable Diffusion AI art tool. The Getty Images lawsuit, filed this week in London’s High Court, alleges that “Stability AI illegally copied and processed millions of copyrighted images,” and used these images for its own commercial gain.
Getty Images and image library rights holders like it are set to be some of the most affected by AI image creation, and while some have embraced the technology, such as Shutterstock (opens in a new tab), others have largely dismissed AI art. Getty Images falls into the latter category. Getty Images banned the uploading and selling of AI images (opens in a new tab) in September 2022, in order to protect itself from legal challenges.
Getty CEO Craig Peters previously said there were concerns about AI-generated artwork, including “unaddressed rights issues.”
Little did we know then, the legal challenges would come directly from Getty herself.
“Getty Images believes that artificial intelligence has the potential to stimulate creative endeavors”, A. Getty Images statement (opens in a new tab) says “Accordingly, Getty Images has provided licenses to leading technological 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 the commercial interests at hand.”
It all comes down to how AI is trained – the most important preliminary step to building a functioning AI tool.
The art created by Stable Diffusion, and other similar tools, is original. There’s probably nothing else quite like it – an AI created it based on whatever prompts it was given by its human users. However, the AI has to be trained on millions of tagged images to learn what to create, and the meaning of the copyright of images used for this training has been questioned on many occasions.
Stability AI did not seek any such license from Getty Images and instead, we believe, chose to ignore viable licensing options and longstanding legal protections.
Getty Images
These large data sets of image and text pairs are often provided by other companies that collect and assemble the data themselves. These are massive datasets, we’re talking millions, or billions, of values for each. This means that collecting enough samples to fill an entire array can be an exhausting job requiring massive amounts of man-made source material – and each and every one of these images has its own copyright position, from public domain to strict copyright or usage terms.
Such a data set organization is Laion (opens in a new tab)which provides the datasets used for Stable Diffusion.
LAION is non-profit and freely distributes its datasets. It basically collects datasets from alt text and image URLs that it is scratches from the internet. He owns nothing. The organization tries to make peace with this in its first question Common questions (opens in a new tab): “Do LAION datasets respect copyright laws?”
To this the organization replies: “LAION datasets are simply indexes to the Internet.”
You might even find a photo of you in LAION’s datasets, for which the organization has provided a GDPR takedown form to maintain compliance with EU law.
LAION tries to keep its hands clean by not actually storing or distributing images directly, and it may have worked because Getty Images’ ire is instead aimed directly at Stability AI, one of the more well-known companies using its datasets.
Stability AI claims that the LAION datasets it uses for Stable Diffusion have been trained in accordance with German law. In response to his frequently asked question “What are the copyrights for using images created in stable diffusion?” Stability AI says “the field of AI-generated images and copyright is complex and will vary from jurisdiction to jurisdiction.”
A really unclear answer.
Stability AI at least says where it gets its data — LAION is actually one of the more open image scraping organizations on the Internet — unlike some AI tools that don’t make that information publicly available. This includes OpenAI (opens in a new tab), the creator of popular AI tools such as DALL-E and ChatGPT. Stability AI plans Allow artists to opt out of Diffusion’s stable image training (opens in a new tab) With future versions, but it seems a bit like the wrong direction. Shouldn’t Stable Diffusion ask the artists permission to use their work?
There is no response to the Stability AI lawsuit at this time, but I have reached out to the company for comment.
The CEO of Getty Images spoke with him the border (opens in a new tab) To further explain why the company took this course of action.
“We do not believe this particular layout of Stability’s commercial offering is covered by fair dealing in the UK or fair use in the US,” Peters said. To protect our intellectual property rights and those of our donors.”
Peters also confirmed that the charges filed against Stability AI include copyright infringement and violation of Getty Images’ terms of service. According to him, the company seeks to create a new legal status quo for licensing and artificial intelligence from the court case.
Such a status quo could be seminal for the nascent AI tools industry. Copyright law still doesn’t catch up with training AI on massive datasets of scraped images, which means the next few years of litigation could shape how we approach this subject for decades to come.
The lawsuit filed by Getty Images could set a legal precedent, whatever the outcome.
It was bound to happen at some point, and the copyright debates surrounding AI-generated images and AI tool training are just beginning.