Supreme Court Opinion Finds Warhol Art Not Protected by Fair Use
Republished from my Art Matters newsletter on Substack
Andy Warhol Foundation has lost a Supreme Court case in a ruling that found the artist violated photographer Lynn Goldsmith’s copyright when he used a photograph of Prince as a reference for some of his artwork. This week’s Supreme Court ruling against the Warhol Foundation will likely have profound legal implications for artists who use appropriated imagery and could reshape the legal doctrine of 'Fair Use' that allows limited use of others’ intellectual property by artists and writers as a means of protecting freedom of expression. It could also impact the work of AI artists who rely on training models which scrape the internet and use art without prior authorization.
Warhol was doing a commission for Vanity Fair in the 1980s when he used the photograph as a reference, and later created a series of other artworks based on the image.
Lynn Goldsmith’s photo of Prince and the artwork created by Andy Warhol
For many years writers and artists have relied on the concept of fair use, which protects them in being able to use text and images in cases where the work is criticism, opinion, journalism and art, where the use is considered to be ‘transformative.’ The justices found that Warhol should not have used the photo as a reference because Goldsmith was also in the business of selling images of Prince, and differentiated between that and the use of say, Campbell’s soup cans. “The purpose of Campbell’s logo is to advertise soup. Warhol’s canvases do not share that purpose,” wrote Supreme Court Justice Sonia Sotomayor.
The court’s opinion, in this case, could apply to rulings in the future, including protecting artists’ and writers’ work from being scraped by AI training models. In his testimony this week before the Senate Judiciary Committee hearing on AI OpenAI CEO Sam Altman told lawmakers that his company would honor requests not to use someone’s data, and he told them that regulation of the burgeoning industry is necessary. You can watch a replay of the hearing here:
In the meantime, I play with the text-to-image robots when I can. I will occasionally invoke the names of some of my favorite non-living artists, such as Wassily Kandinsky, Stuart Davis, or Juan Gris, but I won’t create a prompt based on a living artist. The technology is fascinating. I’m trying to experiment with ways to train AI models on my own artwork, and I’m getting some fascinating results. The AI industry seems to advance almost every day. You can now submit an image along with a text prompt, and it’s kind of a call and response to see what the model generates and then refine the results.
So I’ve submitted my abstract geometric art and included prompts about geometric art and a guitar, and gotten some fascinating results. But if I ask for it to give me a more generic abstract geometric composition, it doesn’t seem to understand how to draw complete, symmetrical geometric forms, or understand how to compose them.
The pixel-based images the AI creates are currently too low of a resolution to be able to be used commercially. But I have used some of the images the AI creates as a reference for work I later create myself. I’ll either use the arrangement of objects the AI creates or use the palette it creates as starting points. Quite a fascinating adventure.
If you would like to support my writing these posts are also on Substack, where I publish the Art Matters website. By the way, the Substack website has created a social media area of the site called Substack Notes which is turning out to be a very interesting alternative to Twitter. The users are well-read and the discussion is very thoughtful and anyone can sign up. If you do, you can read my posts anytime there as well, along with the work of a lot of talented writers. Hope to see you there.
