OpenAI chief executive Sam Altman (Eric Risberg/AP)
OpenAI is placing curbs on its artificial intelligence (AI) image generator after it went viral, unleashing a surge of anime-inspired images based on memes and real-world events, some of which have taken a disturbing turn.
On Thursday, OpenAI CEO Sam Altman acknowledged the overwhelming demand for the tool, saying, "our GPUs are melting", and adding that temporary limits would be imposed while the company worked to improve efficiency.
When OpenAI’s image generator launched earlier this week, it quickly became a sensation, mirroring ChatGPT’s rise to prominence as the go-to AI assistant.
Users jumped on the tool, embracing its ability to create images in a Studio Ghibli-inspired art style, echoing the Oscar-winning Japanese animation studio’s wide-eyed aesthetic.
Studio Ghibli gets twisted
However, the novelty soon gave way to a darker trend. On X, formerly Twitter, some users began circulating provocative and offensive content, including depictions of Hitler, police brutality, and anti-Semitic imagery, with the juxtaposition of such violence against the innocence of Ghibli’s beguiling style lending the images an unsettling visual quality.
The content in question, some of which hails from openly racist accounts, includes one of an orthodox Jewish man chasing money down a street, reinforcing a long-standing and harmful anti-Semitic stereotype.
A separate post depicts the moment Sydney Wilson was fatally shot by a US police officer last September, as captured through a body cam.
Another account, that goes by the name of Sensitive Young Fascist, posted an image of Hitler being cheered on by a crowd, with onlookers giving Nazi salutes.
All of the images are in Studio Ghibli’s signature art style, and together, they have been viewed millions of times.
An alleged AI-generated image depicting the fatal shooting of Sydney Wilson at the hands of a US police officer last September, as captured through a body cam
@mask_bastard / X
OpenAI investigating images for infringements
After The Standard notified OpenAI about the offensive content being generated, the company stated it would review the examples to determine whether they violated its policies.
It also pointed to safeguards in place to block harmful content, such as extremist propaganda and hate speech, adding that users must comply with its policies prohibiting the use of the tool for deception, harassment, or harm.
The company also clarified that its policies specifically forbid repurposing content to defraud, mislead, or promote violence, hatred, or suffering. OpenAI added that before launching new tools, it conducts thorough evaluations, seeks feedback from users and external partners, and implements safety measures to minimise risks.
Read More
Where to buy art in London from cutting-edge galleries to vintage showrooms
Where to buy art in London from cutting-edge galleries to vintage showrooms
Why is everyone posting Studio Ghibli AI images?
Why is everyone posting Studio Ghibli AI images?
Pregnancy scans using AI ‘pick up foetal abnormalities in half the time’
Pregnancy scans using AI ‘pick up foetal abnormalities in half the time’
Sponsored
Past AXA Startup Angel winners share their tips
Past AXA Startup Angel winners share their tips
The Standard has reached out to Studio Ghibli for comment.
Social media users raise ethical concerns
Despite the surge of AI-generated anime images flooding feeds, the broader reaction to OpenAI’s latest tool has been mixed. Some social media users have voiced concerns over issues like copyright, ethical use, and the environmental impact of artificial intelligence.
Christopher Miller, co-director of the acclaimed Spider-Man: Into the Spider-Verse, expressed his disapproval with a simple thumbs-down emoji in response to an AI-generated animation based on the Apple TV+ series Severance.
Others have been more direct in their criticism, highlighting the immense effort and skill required to create original animation, and calling attention to the work of Studio Ghibli filmmaker Hayao Miyazaki, the revered auteur behind award-winning films like Spirited Away, Princess Mononoke and Howl's Moving Castle.
On X, one user expressed their discomfort, saying: “The whole Studio Ghibli AI trend honestly gives me second-hand embarrassment knowing how hard Hayao Miyazaki has fought to retain the identity of his films and how many of you are this willing to make a farce out of decades of artistry because you don’t actually value it."
Another user pointedly added: “Refuse to understand why AI tech bros keep messing around with creative arts instead of focusing on science, engineering, medical or finance. Hope Miyazaki sues them into oblivion. 'Fair use' my a**.”
OpenAI says Studio Ghibli styles are fair game
For its part, OpenAI is pushing ahead with its AI ambitions, even in the face of legal and public challenges. A federal judge recently allowed The New York Times' copyright lawsuit against the company to move forward, in which OpenAI was accused of using the newspaper’s content without permission to train AI models.
Meanwhile, over 400 Hollywood names, including Ben Stiller and Mark Ruffalo, have signed an open letter urging the US government to protect copyright laws in the age of AI.
Seemingly undeterred by the growing tide of opposition, OpenAI has called on the Trump administration to ease access to copyrighted materials, arguing for a need to strengthen America’s position in the global AI race.
On the topic of copyright concerns relating to Studio Ghibli, an OpenAI spokesperson said: “Our goal is to give users as much creative freedom as possible. We continue to prevent generations in the style of individual living artists, but we do permit broader studio styles—which people have used to generate and share some truly delightful and inspired original fan creations.”
Is it just a fad?
As with previous AI experiments, this latest trend exposes a more unsettling phenomenon: the exploitation of AI tools by some users to see how far they can be pushed.
Back in 2016, Microsoft was forced to shut down its Tay chatbot after it was provoked into spouting racist bile. ChatGPT rival, Character AI, is also home to hate-spewing chatbots generated by anonymous users.
Over on X, the AI Ghibli images have already run the full gamut of virality — from novelty to saturation, and now mockery. Still, with the direction of travel in Silicon Valley (and beyond), the next AI-generated spectacle is probably just around the corner.