AI Image Generators Spark Legal Questions Over Studio Styles
OpenAI’s new GPT-4o image generator is taking social media by storm with eerily accurate Studio Ghibli-style art. But the viral fun masks serious legal questions. Experts warn that training AI on copyrighted works—like animated films—might breach copyright law, despite “style” itself not being protected. As courts weigh in, companies like OpenAI and Google continue to push the boundaries of fair use in generative AI.
AI Image Tools Go Viral—But Are They Legal?
It’s only been 24 hours since OpenAI released its native image-generation tool inside ChatGPT, and already the internet looks like it’s been run through a Studio Ghibli filter. Scroll through Twitter, Instagram, or Reddit, and you’ll find anime-style renderings of Elon Musk, Donald Trump, Gandalf, and even Sam Altman, OpenAI’s CEO—each stylized like they stepped out of “Spirited Away.”
The explosion of AI-generated memes feels playful on the surface, but it’s stirring up serious questions beneath. Namely, are OpenAI and Google, which recently released a similar image-generation feature in its Gemini Flash model, walking into legal minefields by enabling users to mimic the unmistakable visual signatures of iconic studios like Ghibli?
The Rise of Studio-Inspired AI Art
Users are feeding photos into ChatGPT’s GPT-4o model and instructing it to “make this in the style of Studio Ghibli” or “recreate this as a Pixar wedding scene.” The results? Eerily accurate imitations of legendary animation styles that, until now, were the exclusive domain of world-renowned creators.
OpenAI appears to have nailed the Ghibli aesthetic better than its rivals. In our informal test against other image generators—Google’s Gemini, xAI’s Grok, and Playground.ai—OpenAI’s version most convincingly echoed the soft textures, muted palettes, and dreamy expressions typical of Hayao Miyazaki’s work.
But that uncanny fidelity is exactly why legal experts are paying attention.
Style Is Not (Yet) Copyrighted — But the Data Might Be
“There’s a legal gray area here,” says Evan Brown, an intellectual property attorney at Neal & McDevitt. Brown explains that while copyright law doesn’t currently protect an artistic style, it does protect the original works that inform that style, like the frames of Ghibli films, potentially used to train the AI.
“If a model was trained on hundreds of thousands of copyrighted images without permission, that’s where the legal problem begins,” Brown told us. “It’s not just about what the model can output. It’s about how it learned to do that.”
This is the heart of ongoing lawsuits from The New York Times, several book authors, and a growing number of artists against companies like OpenAI, Meta, and Midjourney. The plaintiffs argue that using copyrighted data, without payment, credit, or license, to train AI tools violates their intellectual property rights.
OpenAI’s Defense: “Studio Styles” Are Fair Game
In a statement to TechCrunch, an OpenAI spokesperson clarified that ChatGPT is designed to avoid copying the style of individual living artists. However, broader “studio styles”—like those belonging to Pixar or Ghibli—are currently permitted. That’s a nuanced distinction, considering that Ghibli’s style is inseparable from the vision of Hayao Miyazaki, who is alive and remains the studio’s creative backbone.
Even so, OpenAI isn’t alone in walking this legal tightrope. Google’s Gemini model also sparked controversy earlier this month when users discovered it could remove image watermarks—another action with potential copyright implications.
Users Are All In — Courts Are Still Catching Up
The viral popularity of these tools suggests that most users aren’t overly concerned with legal nuance. For them, it’s about fun and creativity—turning a selfie into a Ghibli frame or a vacation shot into a scene from “Up.” One couple even reimagined their wedding album in Pixar’s style using ChatGPT’s image generator.
But behind the fun lies a deeper issue: the legal system is still scrambling to catch up with AI’s capabilities. As courts evaluate whether AI training falls under fair use or constitutes large-scale copyright theft, the stakes are massive, both for creators and the companies building these tools.
For now, OpenAI has delayed access to the image generator for free-tier users due to overwhelming demand, a testament to its popularity. But popularity won’t shield it from scrutiny.
As Brown puts it, “We’re watching the biggest copyright question of our generation unfold in real time.”

