Getty Images has initiated legal proceedings against the creators of AI art tool Stable Diffusion. The Getty Images lawsuit, which was filed in London’s High Court this week, found that “Stability AI illegally copied and processed millions of copyrighted images” and used these images for its own commercial purposes. It claims to have been used for profit.
Getty Images and the rights holders of such image libraries are set to be most impacted by AI image generation. shutter stock (opens in new tab)Getty Images falls into the latter category.Getty Images bans uploading and selling AI images (opens in new tab) September 2022 to protect yourself against legal challenges.
Getty CEO Craig Peters previously said there were concerns about AI-generated artwork, including “unaddressed rights issues.”
Little did we know at the time that the legal challenge would come directly from Getty itself.
“Getty Images believes artificial intelligence has the potential to inspire creative endeavors. Getty Images statement (opens in new tab) Say. “Henceforth, Getty Images has provided licenses to key 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 such a license from Getty Images. I think I did.”
It all depends on how the AI is trained. This is a very important preparatory step for building a functioning artificial intelligence tool.
The art produced by Stable Diffusion and other similar tools is original. The AI was created based on prompts passed by human users. But the AI has to be trained on millions of labeled images in order to learn what to create, and what this means for the copyright of the images used for this training is has been questioned on many occasions.
Stability AI did not seek such a license from Getty Images. Instead, we believe they chose to ignore viable licensing options and longstanding legal protections.
Getty Images
Large datasets of these image-text pairs are often provided by other companies that collect and collate the data. These are large datasets, each with millions or even billions of entries. This means that collecting enough samples to populate the entire dataset can require vast amounts of human-generated source material, with everything from the public domain to strict copyright or usage terms. images have their own copyright.
One such dataset organization is lion (opens in new tab)which provides the dataset used for stable diffusion.
LAION is non-commercial and freely distributes its datasets. It basically matches the dataset from the alt text and image URL. abrasion from the internet. It owns no objects.The organization tries to reconcile it with its first question Frequently Asked Questions (opens in new tab): “Does the LAION dataset respect copyright law?”
The organization responded, “The LAION dataset is simply an index into the Internet.”
You may find your photo in the LAION dataset. LAION offers a GDPR takedown form to comply with EU law.
LAION tries to keep their hands clean by not actually storing or distributing images directly. Getty Images’ ire is directed squarely at Stability AI, one of his high-profile companies using that dataset, so it could have worked.
Stability AI believes that the LAION dataset used for Stable Diffusion was trained in compliance with German law. An answer to its own FAQ question, “What is the copyright for using images generated by Stable Diffusion?” Stability AI states that “AI-generated images and copyright domains are complex and vary by jurisdiction.”
A really vague answer.
Stability AI at least mentions where it gets its data from. LAION is actually one of the more open image scraping organizations on the web. Unlike some AI tools that don’t expose that information.For that Open AI (opens in new tab)is the creator of popular AI tools such as DALL-E and ChatGPT.Stability AI plans Allow artists to opt out of Stable Diffusion image training (opens in new tab) In future versions it seems a bit like the wrong way to go. Shouldn’t Stable Diffusion ask permission to use the artist’s work?
Stability AI has not yet responded to legal action, but we have reached out to the company for comment.
Getty Images CEO said The Barge (opens in new tab) Further explain why the company has taken this course of action.
“We do not believe that this particular deployment of Stability’s commercial offering is covered by fair trade in the UK or fair use in the US,” said Peters. “We did not induce Getty Images to use any material from us or our contributors, and we take steps to protect the intellectual property rights of us and our contributors.”
Peters also confirmed that the charges brought against Stability AI include copyright infringement and violation of Getty Images’ terms of service. He said the company is trying to shape a new legal status quo for licensing and AI out of litigation.
This state of affairs could be formative for the nascent generative AI tools industry. Copyright law has not yet caught up with training AI on massive datasets of scraped images. In other words, the next few years of legal action could shape how we approach this topic for decades to come.
Whatever the outcome, the lawsuit filed by Getty Images is likely to set a legal precedent.
This is bound to happen at some point, and the copyright debate around both AI-generated images and AI tool training is just beginning.
