Tom Hals and Blake Brittain
(Reuters) – Last year, Kris Kashtanova ) fed instructions from a graphic novel into a new artificial intelligence program and sparked a high-stakes debate about who creates the artwork: humans or algorithms.
“Zendaya leaves the gates of Central Park,” Kashtanova enters Midjourney, a ChatGPT-like AI program that generates dazzling illustrations based on written prompts. “Sci-Fi Scenario Future Empty New York…”
From these inputs and hundreds emerge “Zarya of Dawn”, a 21 – The story follows a character similar to actress Zendaya who roams a desolate Manhattan hundreds of years in the future. Kashtanova was awarded the copyright in September, declaring on social media that this means artists have the right to legal protection for their AI art projects.
The good times don’t last long. In February, the U.S. Copyright Office suddenly reversed, and Kashtanova became the first person in the country to be deprived of the legal protection of artificial intelligence art. The images in “Dawn” “are not the product of human creation,” the office said. The office allowed Kashtanova to retain the copyright to the score and storyline.
Now, with the help of a strong legal team, the artist is once again pushing the limits of the law. For a new book, Kashtanova turned to a different AI program, Stable Diffusion, which allows users to scan their own drawings and refine them using text cues. The artist believes that starting with an original work of art will provide enough “human” elements to influence the authorities.
“It would be weird if it wasn’t copyrightable,” the latest work of the 37-year-old artist, an autobiographical comic.
A spokesman for the Copyright Office declined to comment. Midjourney also declined to comment, and Stability AI did not respond to a request for comment. Propagations seem poised to change human expression as they break records for user growth, and the legal system still hasn’t figured out who owns the output — the user, the program owner, or maybe no one at all.
Billions of dollars could depend on the answer, legal experts say. Merkley is the former head of Creative Commons, an organization that issues licenses that allow creators to share their work.
For example, a company could use AI to produce and own the rights to large volumes of low-cost graphics, music, video and text for advertising, branding and entertainment. “Copyright management agencies will be under enormous pressure to grant copyright to computer-generated works,” Merkley said.
In the United States and many other countries, anyone engaging in creative expression generally has direct legal rights. Copyright registration creates a public record of a work and allows owners to go to court to enforce their rights.
Courts, including the U.S. Supreme Court, have long held that the author must be a single person. In denying legal protection to the “Zarya” image, the U.S. Copyright Office cited rulings denying a A selfie taken by a curious monkey named Naruto and a song that the copyright applicant says was composed by the “holy spirit” offer legal protection.
Stephen Thaler, an American computer scientist from Missouri, insists that his AI programs are sentient and should be legally recognized as creators of the artwork and inventions they produce. He has sued the US Copyright Office, petitioned the US Supreme Court, and tried patent cases in the UK Supreme Court.
At the same time, many artists and companies with creative content strongly oppose granting copyright to AI owners or users. They argue that because the new algorithm works by training itself on vast amounts of material on the open web, some of which is copyrighted, the AI system is gobbling up legally protected material without permission.
Stock photo provider Getty Images, a group of visual artists, and the owner of computer code have each filed copyright infringement lawsuits against owners of AI programs including Midjourney, Stability AI, and ChatGPT developer OpenAI , these companies deny. Getty and OpenAI declined to comment.
Sarah Andersen, one of the artists, said that copyrighting AI’s work “would legitimize theft”.
“The Problem”
Kashtanova was represented pro bono by Morrison Foerster and his senior copyright attorney Joe Gratz, who also defended OpenAI in the proposed class action Owner of protected computer code. The firm took on Kashtanova’s case after a partner at the firm, Heather Whitney, discovered a LinkedIn post of the artist seeking legal help with a new application after the “Zarya” copyright was rejected.
“These are hard questions that have a big impact on all of us,” Gratz said. Posting on Instagram, the company said the images were created using AI, saying it wasn’t clear in the original filing in September. In March , it issued public guidelines requiring applicants to explicitly disclose whether their work was created with the help of artificial intelligence.
The most popular AI systems may not create copyrighted works, the guidance says, “what matters is the degree of creative control that humans have.”
“Totally bragging”
Kashtanova, who identifies as non-binary and uses “they/them” pronouns, discovered Midjourney in August after the pandemic shut down their jobs largely because of yoga static Photographer of repairs and extreme sports events.
“I was completely blown away,” the artist said. Now, with AI technology advancing at lightning speed, Kashtanova is turning to newer tools that allow users to input original compositions and give more specific commands to control the output.
To test the extent to which human control can satisfy the Copyright Office, Kashtanova plans to file a series of copyright applications for individual images selected from the new autobiographical comic, each imaged in a different Made by AI programs, settings or methods.
The artist, who now works at a startup that uses AI to turn children’s drawings into comic books, created the first such image a few weeks ago, titled “Rose Enigma.”
Kashtanova sat down at their computer in their one-bedroom Manhattan apartment to demonstrate their latest technology: They pulled up a simple pen-and-paper sketch on the screen, which they had scanned into Stable Diffusion , and then began to refine it by tweaking settings and using text cues, such as “young robot woman” and “flowers sprouting from her head.”
The result is an otherworldly image, with the lower half of a woman’s face replaced by a long-stemmed rose for the upper half of her head. Kashtanova filed for copyright protection 16 in March.
This image will also appear in Kashtanova’s new book. It’s titled: “For My AI Community”.
