This artist is dominating AI-generated art. And he’s not happy about it.
Those cool AI-generated images you’ve seen across the internet? They are likely to be based on the Greg Rutkowski . works.
Rutkowski, a Polish digital artist, uses classical painting techniques to create dreamy fantasy environments. He has illustrated games like Sony’s Horizon Forbidden West Ubisoft’s Anno , Dungeons and Dragons ,, and Magic The Gathering . He’s a huge hit in the new world text-to-image Ai generation.
His distinctive style is now one of the most commonly used prompts in the new open-source AI art generator Stable Diffusion, which was launched late last month. This tool, along other popular image-generation AI models allows anyone to create stunning images based upon text prompts. Type in “Wizard with a sword and a glowing orb o magic fire fights a fierce Dragon Greg Rutkowski” and the system will generate something that is not far from Rutkowksi’s style.
But these open-source programs are built by scraping images from the internet, often without permission and proper attribution to artists. These programs raise questions about copyright and ethics. Artists like Rutkowski are fed up.
According to the website Lexica, which tracks over 10 million images and prompts generated by Stable Diffusion, Rutkowski’s name has been used as a prompt around 93,000 times. Some of the world’s most famous artists, such as Michelangelo, Pablo Picasso, and Leonardo da Vinci, brought up around 2,000 prompts each or less. Rutkowski’s name is also featured as a prompt thousands times in the Discord of Midjourney, an image-to-text generator.
Rutkowski was initially surprised but thought it might be a good way to reach new audiences. He tried searching his name to see if any of the pieces he had written had been published. He was able to find work that had his name attached but not his.
It’s been a month. What about in a year. Rutkowski states that it is unlikely that I will be able find my work online because the internet will be overloaded with AI art. “That’s concerning.”
Stability.AI, the company that built Stable Diffusion, trained the model on the LAION-5B data set, which was compiled by the German nonprofit LAION. LAION put the data set together and narrowed it down by filtering out watermarked images and those that were not aesthetic, such as images of logos, says Andy Baio, a technologist and writer who downloaded and analyzed some of Stable Diffusion’s data. Baio analyzed 12 million of the 600 million images used to train the model and found that a large chunk of them come from third-party websites such as Pinterest and art shopping sites such as Fine Art America.
Many of Rutkowski’s artworks were taken from ArtStation, where many artists upload their online portfolios. There are many reasons Rutkowski is so popular as an AI prompt.
First, his fantastical and ethereal style looks very cool. His illustrations are also very prolific and are easily accessible online in good enough quality. An early text-to-image generator called Disco Diffusion offered Rutkowski as an example prompt.
Rutkowski also added alt text in English to his online work. These descriptions are helpful for people who use screen readers software and help search engines rank the images. These images are easy to scrape and the AI model can identify which images are relevant to prompts.
Stability.AI released the model into the wild for free and allows anyone to use it for commercial or noncommercial purposes, although Tom Mason, the chief technology officer of Stability.AI, says Stable Diffusion’s license agreement explicitly bans people from using the model or its derivatives in a way that breaks any laws or regulations. The users are responsible for this.
Some artists may have been harmed in the process
Other artists besides Rutkowski have been surprised by the apparent popularity of their work in text-to-image generators–and some are now fighting back. Karla Ortiz is an illustrator from San Francisco who found her work through Stable Diffusion’s database set. She has been raising awareness about copyright and AI art.
Artists fear losing their income if they use AI-generated images that are based on copyrighted materials for commercial purposes. Ortiz says that art is more personal than it seems, and that it could cause privacy and data protection problems. “There is a growing coalition within the artist industries to find ways to mitigate or tackle this,” Ortiz says. This group is still in its infancy and could push for new regulations or policies.
One suggestion is that AI models could be trained on images in the public domain, and AI companies could forge partnerships with museums and artists, Ortiz says.
” “It’s not only artists… It is photographers, models, actors, actresses and directors, cinematographers, and directors, too,” she said. “Any kind of visual professional is dealing with this particular question right now.” Currently, artists have no choice but to opt in or have their work removed from the database. Carolyn Henderson, the manager for her artist husband, Steve Henderson, whose work was also in the database, said she had emailed Stability.AI to ask for her husband’s work to be removed, but the request was “neither acknowledged nor answered.”
“Open-source AI is a tremendous innovation, and we appreciate that there are open questions and differing legal opinions. They will be resolved as AI becomes more widespread and different groups reach a consensus on how to balance individual rights with essential AI/ML research,” Mason, Stability.AI’s spokesperson, said. “We try to find the right balance between helping the community and innovating.
Rutkowski’s “Castle Defense, 2018” (left) and a Stable Diffusion prompted image.
Mason encourages any artists who don’t want their works in the data set to contact LAION, which is an independent entity from the startup. LAION did not respond to a request for comment immediately.
Berlin-based artists Holly Herndon and Mat Dryhurst are working on tools to help artists opt out of being in training data sets. They launched a site called Have I Been Trained, which lets artists search to see whether their work is among the 5.8 billion images in the data set that was used to train Stable Diffusion and Midjourney. Some online art communities, such as Newgrounds, are already taking a stand and have explicitly banned AI-generated images.
An industry initiative called Content Authenticity Initiative, which includes the likes of Adobe, Nikon, and the New York Times, are developing an open standard that would create a sort of watermark on digital content to prove its authenticity. It could be used to combat disinformation and ensure digital creators receive proper attribution.
It could also allow IP holders or creators to claim ownership over media that they own or synthesized media created with their trademark,” Nina Schick, an expert in deepfakes or synthetic media, says.
AI-generated art poses tricky legal questions. According to Gill Dennis, a lawyer at Pinsent Masons, scraping images from the web without permission could be a violation of copyright. This is in the UK, where Stability.AI’s headquarters are located. You can use copyrighted works to train an AI, but only for noncommercial purposes. Stable Diffusion can be used for free, but Stability.AI sells premium access through a platform called DreamStudio.
The UK, which hopes to boost domestic AI development, wants to change laws to give AI developers greater access to copyrighted data. These changes would allow developers to access copyrighted works to train their AI systems commercially and noncommercially.
While artists and other rights holders would not be able to opt out of this regime, they will be able to choose where they make their works available. The risk is that the art community will move into a subscription or pay-per-play model similar to the one used in film and music.
“The risk is that rights holders refuse to make their works public, which would undermine the reason for fair use in the AI space.” Dennis says.
In the US, LinkedIn lost a case in an appeals court, which ruled last spring that scraping publicly available data from sources on the internet is not a violation of the Computer Fraud and Abuse Act. Google also won a case against authors who objected to the company’s scraping their copyrighted works for Google Books. Rutkowski said he doesn’t blame people who use the prompt Rutkowski. He says that it’s a “cool experiment” for them. “But for me and many other artists, it’s starting to look like a threat to our careers.”
I’m a journalist who specializes in investigative reporting and writing. I have written for the New York Times and other publications.