ChatGPT Turned Into a Studio Ghibli Machine. How Is That Legal?

Yahoo

13-05-2025

A few weeks ago, OpenAI pulled off one of the greatest corporate promotions in recent memory. Whereas the initial launch of ChatGPT, back in 2022, was 'one of the craziest viral moments i'd ever seen,' CEO Sam Altman wrote on social media, the response to a new upgrade was, in his words, 'biblical': 1 million users supposedly signed up to use the chatbot in just one hour, Altman reported, thanks to a new, more permissive image-generating capability that could imitate the styles of various art and design studios. Altman called it 'a new high-water mark for us in allowing creative freedom.'
Almost immediately, images began to flood the internet. The most popular style, by a long shot, was that of Studio Ghibli, the Japanese animation studio co-founded by Hayao Miyazaki and widely beloved for films such as Spirited Away and Princess Mononoke. Ghibli's style was applied to family portraits, historical events including 9/11, and whatever else people desired. Altman even changed his X avatar to what appears to be a Ghiblified version of himself, and posted a joke about the style's sudden popularity overtaking his previous, supposedly more important work.
The Ghibli AI phenomenon is often portrayed as organic, driven by the inspiration of ChatGPT users. On X, the person credited with jump-starting the trend noted that OpenAI had been 'incredibly fortunate' that 'the positive vibes of ghibli was the first viral use of their model and not some awful deepfake nonsense.' But Altman did not appear to think it was luck. He responded, 'Believe it or not we put a lot of thought into the initial examples we show when we introduce new technology.' He has personally reposted numerous Ghiblified images in addition to the profile picture that appears atop every one of his posts, which he added less than 24 hours after the Ghibli-esque visuals became popular; OpenAI President Greg Brockman has also recirculated and celebrated these images.
[Read: Generative AI is challenging a 234-year-old law]
This is different from other image-sharing trends involving memes or GIFs. The technology has given ChatGPT users control over the visual languages that artists have honed over the course of their careers, potentially devaluing those artists' styles and destroying their ability to charge money for their work. Existing laws do not explicitly address generative AI, but there are plausible arguments that OpenAI is in the wrong and could be liable for millions of dollars in damages—some of those arguments are now being tested in a case against another image-generating AI company, Midjourney.
It's worth noting that OpenAI and Studio Ghibli could conceivably have a deal for the promotion, similar to the ones the tech company has struck with many media publishers, including The Atlantic. But based on Miyazaki's clear preference for hand-drawn work and distaste for at least certain types of computer-generated imagery, this seems unlikely. Neither company answered my questions about whether such a deal had been made, and neither Miyazaki nor Studio Ghibli have made any public remarks on the situation.
Individual works of art are protected by copyright, but visual styles, such as Studio Ghibli's, are not. The legal logic here is that styles should be allowed to evolve through influence and reinterpretation by other artists. That creative and social process is how van Gogh led to Picasso, and Spenser to Shakespeare. But a deluge of people applying Ghibli's style like an Instagram filter, without adding any genuine creative value, isn't a collective effort to advance our visual culture. The images are also the direct result of a private company promoting a tech product, in part through its executives' social media, with the ability to manufacture images in a specific style. In response to a broader request for comment, a spokesperson for OpenAI told me, 'We continue to prevent generations in the style of individual living artists, but we do permit broader studio styles.'
Still, this has the flavor of an endorsement deal, such as the ones Nike has made with LeBron James, and Pepsi with Beyoncé: Use ChatGPT; make Studio Ghibli art! These kinds of endorsements typically cost millions of dollars. Consider what happened in 1985, when Ford Motor Company wanted to promote one of its cars with an ad campaign featuring popular singers. Ford's advertising agency, Young & Rubicam, asked Bette Midler to record her hit song 'Do You Want to Dance?' but she declined. Undeterred, they approached one of Midler's backup singers and asked her to perform the song in Midler's style. She accepted, and imitated Midler as well as she could. The ad aired. Midler sued.
In court, the judge described the central issue as 'an appropriation of the attributes of one's identity,' quoting from a previous case that had set precedent. Young & Rubicam had chosen Midler not because they wanted just any good singer but because they wanted to associate their brand with the feelings evoked by Midler's particular, recognizable voice. 'When a distinctive voice of a professional singer is widely known and is deliberately imitated in order to sell a product,' wrote the court, 'the sellers have appropriated what is not theirs.' Young & Rubicam had violated Midler's 'right of publicity,' in the language of the law. Midler received a $400,000 judgment (the equivalent of approximately $1 million today).
[Read: The unbelievable scale of AI's pirated-books problem]
OpenAI risked ending up in a similar lawsuit last year when it used a voice many people thought sounded similar to Scarlett Johansson's to promote its voice-assistant product. Like Midler, Johansson had been asked to participate, and declined. Experts believed she had a viable right-of-publicity case against OpenAI. Johansson's lawyers sent letters to OpenAI but did not file a formal legal complaint. (OpenAI denied that the voice was modeled on Johansson's, but removed it and apologized to the actor.)
The average person seeing a torrent of images in the Studio Ghibli style, with captions praising ChatGPT, might reasonably infer that Miyazaki himself endorses or is associated with OpenAI, given that he is the most famous artist at the studio and has directed more of its films than any other. That people tend to call the aesthetic Ghibli's doesn't change the fact that the style is most recognizably Miyazaki's, present even in his early work, such as the 1979 film Lupin III: The Castle of Cagliostro, which was created six years before Ghibli was founded. Surely many people recognize Spirited Away as Miyazaki's and have never heard of Studio Ghibli.
Besides a right-of-publicity complaint, another legal option might be to file a complaint for false endorsement or trade-dress infringement, as other artists have recently done against AI companies. False endorsement aims to prevent consumer confusion about whether a person or company endorses a product or service. Trade-dress law protects the unique visual cues that indicate the source of a product and distinguish it from others. The classic Coca-Cola bottle shape is protected by trade dress. Apple has also acquired trade-dress protection on the iPhone's general rectangular-with-rounded-corners shape—a design arguably less distinctive (and therefore less protectable) than Ghibli's style.
In August, a judge agreed that false-endorsement and trade-dress claims against Midjourney were viable enough to litigate, and found it plausible that, as the plaintiffs allege, Midjourney and similar AI tools use a component that functions as 'a trade dress database.'
[Read: There's no longer any doubt that Hollywood writing is powering AI]
Regardless of what the courts decide or any action that Studio Ghibli takes, the potential downsides are clear. As Greg Rutkowski, one of the artists involved in the case against Midjourney, has observed, AI-generated images in his style, captioned with his name, may soon overwhelm his actual art online, causing 'confusion for people who are discovering my works.' And as a former general counsel for Adobe, Dana Rao, commented to The Verge last year, 'People are going to lose some of their economic livelihood because of style appropriation.' Current laws may not be up to the task of handling generative AI, Rao suggested: 'We're probably going to need a new right here to protect people.' That's not just because artists need to make a living, but because we need our visual aesthetics to evolve. Artists such as Miyazaki move the culture forward by spending their careers paying attention to the world and honing a style that resonates. Generative AI can only imitate past styles, thus minimizing the incentives for humans to create new ones. Even if Ghibli has a deal with OpenAI, ChatGPT allows users to mimic any number of distinct studio styles: DreamWorks Animation, Pixar, Madhouse, Sunrise, and so on. As one designer recently posted, 'Nobody is ever crafting an aesthetic over decades again, and no market will exist to support those who try it.'
Years from now, looking back on this AI boom, OpenAI could turn out to be less important for its technology than for playing the role of provocateur. With its clever products, the company has rapidly encouraged new use cases for image and text generation, testing what society will accept legally, ethically, and socially. Complaints have been filed recently by many publishers whose brands are being attached to articles invented or modified by chatbots (which is another kind of misleading endorsement). These publishers, one of which is The Atlantic, are suing various AI companies for trademark dilution and trademark infringement, among other things. Meanwhile, as of today, Altman is still posting under his smiling, synthetic avatar.
Article originally published at The Atlantic