You’ve probably seen the impressive and weird products of Dall-E, Midjourney, and Stable Diffusion, so-called AI image creators that generate images from a soup of existing works, gathered from the internet. So far, AI art has been a curiosity, a tool for nerdy artists, and a fun way to waste some time. But now, apps that charge money to generate images based on those copyrighted sources without permission are appearing. Things are about to get ugly.  “This issue is that AI’ art’ is not actually ‘art.’ It wasn’t created from the imagination of a human. It was, instead, pieced together from countless already existing art pieces. It’s a more complicated paint-by-numbers scheme. The more consumers use these apps and support AI instead of real, actual artists, the closer we will be to the death of real art. It’s going to happen piece by piece, and it will be impossible to stop because proving that your art was stolen to generate new art is going to be next to impossible,” artist and graphic designer Amy Weiher told Lifewire via email. 

AI Is Not Intelligent

To use one of these image tools, you type a text prompt describing the image you want, and it creates it for you. It can do this because it has been trained on millions of existing images from the internet. If you ask for a picture of a 1970s kids’ TV show puppet, it doesn’t know what a puppet is, or the 1970s, or even a TV show. But by scraping all those human-created works and noting the text and descriptions around them, it can serve up something similar.  The problem here is that many, probably most, of those source works are copyrighted, and the artists have not given permission for them to be used. This appropriation is hidden behind the buzzword ‘AI,’ although there is no more intelligence here than in your computer’s spell check.  That was fine when Open AI (the company behind Dall-E and Chat GPT) and others were just doing research. But as soon as these tools are used commercially, we’re solidly into the realm of copyright infringement.  Type something like “sunflower oil painting Van Gogh,” and you’ll get images that are based very closely on the work of the ear-slicing Dutch artist. That’s fine for long-dead artists, but for living artists, you have used their work without permission. 

Now that money is being made from copyrighted works, big guns like Marvel and Disney might step in and start suing, which could end up being the best thing for everyone. After all, there’s zero chance that any individual will be able to stop their images from being sucked in by the web-scraping AI behemoths.  “The chances will be much better if large companies, especially major copyright holders like Disney, get involved. Individual artists aren’t likely to have the leverage to change anything in this space,” Ben Michael, attorney at Michael and Associates, told Lifewire via email. But even that might not be enough in this case. Google has already been through the courts regarding scanning books and other printed material. The use for that material might be different (archival, search, and presumably advertising), but if commercial companies start ingesting images, it could be relevant.  “The precedent was set long ago, in Authors Guild v Google: which ruled that a machine reading your code/book/movie transcript isn’t copyright infringement unless you’re publishing a non-trivial reproduction of someone else’s code. Some may argue that Copilot [a programming AI helper] copying a few lines of code is trivial, but that will be up to the courts to decide,” Viputheshwar Sitaraman, public speaker on AI, creativity, and automation, told Lifewire via email. It’s a mess, to be sure, and if the recent past is anything to go by, one that isn’t likely to end well for individual artists. AI selfie filters may turn out to be a fad, but the tools for generating images, text, and videos based on the work of others are becoming established. The time to clarify the law around that is right now, but again, looking at the history of the internet, that’s unlikely to happen until it’s too late.