Since March 2025, OpenAI’s upgraded GPT-4o model has revolutionized digital creativity by enabling users to transform photos into Studio Ghibli-inspired art effortlessly. This surge in AI-generated art has led to a cultural phenomenon on social media, raising concerns about the legality of using copyrighted styles for training. While the aesthetic replicating isn’t inherently illegal, questions arise regarding potential copyright infringement if original works were used without permission. The legal landscape remains complex and evolving as users embrace this creative technology.
The Rise of AI-Generated Art
Since March 26, 2025, the online world has experienced an explosion of creativity. OpenAI has upgraded its GPT-4o model, introducing a groundbreaking image rendering engine that can turn any photograph into a masterpiece inspired by Studio Ghibli, the renowned Japanese animation studio behind classics like *My Neighbor Totoro* and *Spirited Away*.
How It Works
Now, you can effortlessly generate images using ChatGPT and Sora, which provide a seamless and free experience. The results? Social media platforms are inundated with imaginative memes featuring figures like Elon Musk as a Ghibli character, Donald Trump as a mystical forest spirit, and Sam Altman, the CEO of OpenAI, showcasing a reimagined portrait as his profile picture. However, this rapid rise in popularity has led to challenges, with OpenAI’s servers struggling to keep pace with user demand, raising some legal concerns.
So, how does this innovative tool operate? Simply upload a photo to ChatGPT, request the model to recreate it in the Ghibli aesthetic, and within moments, you’ll receive an animated version reminiscent of a Hayao Miyazaki film.
No artistic expertise is necessary: the user-friendly interface yields stunning results. Earlier this month, Google attempted something similar with its Gemini Flash model, allowing users to experiment with various styles and even remove watermarks. However, OpenAI has taken performance to new heights, making a significant impact.
The Meme Phenomenon
This phenomenon is more than just a fleeting trend; it’s a genuine cultural wave. In just one day, social media feeds across platforms like X, Threads, Bluesky, Instagram, and TikTok were filled with images produced by GPT-4o. The tool’s effectiveness has made it a meme in its own right, with users competing in creativity and sharing content widely. However, this immense success has prompted OpenAI to limit access for free users due to overwhelming demand.
What truly sets this technology apart is its capability to mimic intricate artistic styles. The “rendering engine”—a technical term for the AI component that generates images—has been trained on vast amounts of visual data, enabling it to recognize the soft features, pastel colors, and ethereal ambiance characteristic of Ghibli films. But this raises a crucial question: where do these training data originate? If OpenAI utilized images from the studio’s films without permission to train its AI, it could lead to significant legal issues.
Legal Implications of AI Art Generation
Amid the laughter and creativity, a pressing question arises: is this legal? The stylistic elements—let’s say “Ghibli”—are not covered by copyright. In essence, replicating the aesthetic of a studio or artist isn’t inherently illegal. However, if OpenAI trained its model on millions of images from Miyazaki’s films without authorization, the situation becomes complicated, entering a legal “gray area.”
This debate is not new. Various media organizations, including major publications, have already taken legal action against OpenAI, alleging that the company has used their content to train its models without consent or compensation. Other AI companies like Midjourney and Meta are also facing scrutiny for similar reasons. OpenAI asserts that it doesn’t copy the style but instead reproduces “broader studio styles.” Yet, the distinct visual identity shaped by figures like Hayao Miyazaki complicates matters. Where exactly is the line drawn?
In France, the legality of using an AI like OpenAI’s to produce images in the style of protected works, such as those created by Studio Ghibli, is equally murky. Copyright law in France protects original creations as soon as they are expressed, covering films, illustrations, and the unique styles of artists like Hayao Miyazaki. This law grants creators (or their rights holders) exclusive rights over the reproduction or adaptation of their work. Thus, if an AI directly employs images or excerpts from Ghibli films for training and production, it could lead to copyright infringement, especially if done without the necessary permissions.
However, the complexity increases as French and European law does not explicitly declare whether training an AI on protected works is illegal in itself. The European copyright directive of 2019 introduces an exception known as “text and data mining,” allowing for the analysis of content—even protected material—to extract information, as long as rights holders have not explicitly forbidden this use. In France, this provision is included in the Intellectual Property Code (article L. 122-5-3). Thus, if Studio Ghibli or its representatives prohibit the use of their works for AI training, it could halt progress.
As for using the generated images (like transforming a photo into Ghibli style), the situation differs. If it’s done for personal enjoyment and kept private, it often falls under the “private copy” exception—legal as long as it’s not shared publicly. However, sharing on social media or selling these images enters a risky territory, as rights holders might view it as commercial exploitation of their style without consent, even if the AI has “recreated” the image from scratch. While the style itself isn’t directly protected, if it’s too recognizable and associated with a specific work, French courts could interpret it as infringement or unfair competition.
Currently, the courts have yet to make definitive rulings. Some judges are contemplating whether using protected works to train an AI could be classified under “fair use,” a legal exception allowing certain uses without permission. In the meantime, users are reveling in this creative freedom, and it’s easy to see why.