A few days ago, The OpenAI company He launched a new image update in Chatgpt, in which, according to its executive director, Sam Altman, “is an incredible technology/product.”
Users noticed that, With this update, Chatgpt allowed generating images with styles of different animation studies, including – and the most popular— Studio Ghibli.
The curiosity of many led them to try this option and, since then, social networks began to flood photos converted into illustrations in the style of My neighbor Totoro either Chihiro’s trip.
However, this option, in addition to unleashing a wave of creativity in social networks, also created a debate on the limits of artificial intelligence (AI) against intellectual property.
The trend became so viral that Altman himself and some members of the Openai team changed their profile photo in the social network X for a Ghibli style version of themselves.
We are launching to New Thing Today – Images in Chatgpt!
Two Things To Say About It:
1. it’s an incredible Technology/Product. I remember seeing sum of the first images eat of this model and haVing a Hard Time they were Really Made by ai. We Think People Will Love It, and We …
– Sam Altman (@sama) March 25, 2025
How did the creation of the images work?
Those who managed to use the tool before the restrictions followed these steps:
- Users accessed chatgpt.
- They loaded their own or even taken from the Internet (preferably with visible faces and good lighting).
- They provided a simple instruction, as “convert this photograph in the style of Studio Ghibli.”
- They received an image with anime aesthetics, vibrant colors and cartolescos faces.
Even so, the platform did not always obey: In some cases, I rejected orders that included direct references to the study.
Shortly after, Some users began receiving error messages when trying to replicate the style, with warnings on copyright. Although Openai has not confirmed an explicit block, the measure seems to respond to the thin line between the inspiration and appropriation of a protected artistic seal.
However, people took advantage of this function to recreate memes or animated versions of historical moments, pay tributes or simply convert one of their photographs with this style.
Studio Ghibli style guatemala thread pic.twitter.com/31H3HE33YQ
– Geo d’Antu 🦕 (@yosh_g) March 28, 2025
The legal dilemma: Can an AI imitate an artistic style?
Some years ago, Hayao Miyazaki, director of Studio Ghibli, in the documentary Hayao Miyazaki: The One Who Never Ends, He described the art generated by algorithms as “an offense to life itself”, defending artisanal work and authenticity in animation.
Miyazaki has publicly rejected AI and, at that time, he said: “Whoever believes this,” taking you generated with AI, “has no idea of pain or anything. I am deeply disgusting. If you really want to do disgusting things, continue with it. I would never want to incorporate this type of technology into my work. I feel deeply that this is an insult to life itself.” For this reason, the company prohibits imitating styles of living artists.
However, among the conflicts that could generate the creation of this type of images, are:
- Copyright: Although a “style” cannot be legally protected, specific elements (such as Ghibli characters) are registered.
- Creative Ethics: Among the debates, the questioning of whether it is fair that an AI replicates artists who dedicated decades to perfect them has arisen.
- Consent: The AI models are trained with millions of images, many without permission from their creators.
What does Openai say on the subject?
The company has clarified that Its tools should not be used to violate copyright, but avoid mentioning specific cases such as Ghibli. Joanne Jang, one of the company’s behavior model, shared a statement on her social networks, which was repossed by the official OpenAI account.
In this statement, he said: “IA laboratory employees should not be the referees of what people should and should not create. We always feel humble after launch, when discovering cases of use that we never imagine, or even cases that seem so obvious in retrospect, but that did not occur to us from our limited perspectives. ”
Jang ended the text by saying: “The future is built with imagination and adventure. As long as we continue to investigate and learn from society, we believe that we can continue to find ways to responsibly increase the freedom of users. When our policies evolve (not if they evolve!), Update them based on the comments of the real world is not a failure; that is the objective of the iterative implementation.” Besides, He encouraged users to continue sharing their comments and creations.
// i lead model Behavior at Openai, and Wted to Share Sub Thoughts & Nuance That Went into Setting Policy for 4o Image Generation.
features capital letters (!) Bc i published it as a blog post:
–
This Week, We Launched Native Image Generation in Chatgpt Through 4O.It was …
– Joanne Jang (@joannejang) March 27, 2025
Meanwhile, the controversy reflects a cultural shock: on the one hand, there are those who celebrate the democratization of art; on the other, those who see in AI a threat to originality. The truth is that, as these technologies advance, the legislation will have to define how far the “inspiration” comes and where the violation begins.