On March 25, 2025, OpenAI announced an update to ChatGPT’s image generation capabilities, integrating the GPT-4o model (which until now was limited to generating text).
This new feature allows users to create visuals directly from their conversations, eliminating the need to use external tools. This advance is part of the universe of OpenAI solutions for the user.
This innovation has given rise to a viral trend that has captured global attention: the transformation of images in the style of Studio Ghibli, the legendary Japanese animation studio known for masterpieces such as My Neighbor Totoro or Spirited Away.
It has also generated intense controversy over copyright, ethics, and the impact of artificial intelligence (AI) on art. Let’s explore what has happened.
New image generation features in ChatGPT
The arrival of the GPT-4o model marks a before and after in ChatGPT’s capabilities for image generation.
This model, which replaces the previous DALL-E 3, is initially available to Pro and Plus subscribers, although OpenAI plans to progressively extend it to free plan users and developers through its API.
Among the most notable improvements is the ability to generate more precise and detailed images, as well as edit existing images, including photos with people. For example, users can add objects in the foreground or background, adjust colors, or completely transform an image using textual instructions.
Is GPT-4o better than DALL-E 3?
Technically, GPT-4o is now a multimodal model that combines text and images more seamlessly than its predecessors.
While DALL-E 3 was primarily limited to creating images from textual prompts, GPT-4o allows for iterative interaction: users can modify their creations step by step, refining details with unprecedented precision
This functionality is based on extensive training with public and proprietary data, including images licensed from Shutterstock, reflecting an effort by OpenAI to respect creators’ rights.
These capabilities open up a range of possibilities for creating personalized visual content. Ý also for amateurs who experiment with editing their personal photos, as happened almost immediately.
Studio Ghibli’s viral trend
Shortly after the release of this update, ChatGPT users began exploring a creative application that quickly went viral: transforming images in the iconic Studio Ghibli style.
This studio, founded by Hayao Miyazaki and Isao Takahata, is renowned for its unique aesthetic, combining soft color palettes, lush backgrounds, and deeply emotional storytelling.
Movies like Princess Mononoke and Howl’s Moving Castle have set a standard for excellence in animation, and now, thanks to AI, anyone can recreate that magic.
On platforms like X and Instagram, images generated in the Ghibli style have flooded the networks. From photos of pets turned into animated characters to reinterpretations of historical moments or even memes transformed into bucolic scenes, the creativity of users has no limits.
Even Sam Altman, CEO of OpenAI, joined the trend by changing his X profile photo to a Ghibli-style version of himself, TechCrunch reported in 2025.
The popularity has been so overwhelming that OpenAI has had to delay access to this feature for free users due to high demand, which has strained the capacity of its servers.
Studio Ghibli evokes childhood memories for millions of people, and the possibility of integrating that style into everyday life through AI has resonated deeply with audiences. However, this phenomenon has not been exempt from criticism and questioning.
The controversy over Copyright
The use of ChatGPT to replicate Studio Ghibli’s style has sparked a hot debate over copyright and intellectual property.
Although the style itself is not protected by copyright, there is a suspicion that GPT-4o has been trained with millions of frames from Studio Ghibli films, material that is protected.
This puts the technology in a legal gray area: is training AI models with protected works a infringement or does it fall under the fair use doctrine? The courts still do not have a definitive answer.
OpenAI has attempted to mitigate these concerns by stating that ChatGPT will not replicate the style of individual living artists, allowing only the emulation of broader styles such as that of Studio Ghibli.
However, this distinction is questionable, as the studio’s style is intrinsically linked to the vision of creators like Hayao Miyazaki, who is still alive and active.
Recent legal cases in the US, such as Andersen v. Stability AI, have set important precedents.
This judgment determined that both companies developing AI and end users could be liable for infringement if unlicensed data is used or if the generated images violate existing rights.
This scenario highlights the urgent need for a legal framework that regulates the use of AI in the creation of visual content, an area where technology is advancing much faster than legislation.
“An insult to life itself” says Miyazaki
Beyond the legal, the trend raises profound ethical dilemmas. Hayao Miyazaki has been especially critical of AI-generated art, describing it as “an insult to life itself” and questioning its creative value and authenticity.
For Miyazaki and other artists, AI represents a potential threat to the livelihoods of illustrators and animators, especially in a context where the Studio Ghibli style is a symbol of artisanal excellence.
The use of AI to imitate styles without consent can also be interpreted as a form of cultural appropriation, particularly when generated images go viral without acknowledging the original source.
This also raises questions about the balance between technological innovation and respect for human creativity. Is it ethical for a machine to replicate a studio’s decades of work without its approval?
The future of this technology will depend on how these issues are addressed. It is imperative to develop a legal and ethical framework that allows AI to flourish without undermining the rights of creators or the authenticity of art.
This post is also available in: