- OpenAI’s new image-generation tool is so popular it’s “melting” the company’s GPUs, forcing temporary limits, according to CEO Sam Altman. The tool has gone viral for its ability to create Studio Ghibli-style AI, reigniting copyright concerns.
OpenAI has temporarily imposed a rate limit on ChatGPT’s image generation due to overwhelming demand, CEO Sam Altman said, adding that the surge in usage has strained the company’s GPUs.
“It’s super fun seeing people love images in ChatGPT. But our GPUs are melting. We are going to temporarily introduce some rate limits while we work on making it more efficient,” he said in a post on X. “Hopefully won’t be long! ChatGPT free tier will get 3 generations per day soon.”
Altman didn’t specify what the new rate limit is, but the move highlights the significant computing power required for AI-generated images.
Generative AI is highly compute-intensive due to the vast number of calculations needed to process and generate outputs. Training and running these models also require powerful GPUs, large-scale cloud infrastructure, and significant energy consumption.
OpenAI’s native image-generating
The company launched the native image-generating feature, powered by the GPT-4o model, on Tuesday. The new tool produces more realistic images and improves text rendering, with Altman calling it a “new high-water mark for us in allowing creative freedom.”
However, users have mainly been using it to create AI-generated memes in the style of Studio Ghibli, the Japanese animation studio behind films such as My Neighbor Totoro and Spirited Away. Over the last few days, social media feeds have been flooded with Ghibli-versions of people, animals, historical events, and podcasts.
Even Altman has joined in the trend, updating his X display picture to a Ghibli version of himself.
However, the flood of Ghibli-style images has raised questions about potential copyright violations. Many artists have taken issue with AI image generators, arguing that vast datasets used to train models contain copyrighted works without explicit permission from creators. OpenAI is in the middle of fighting several copyright lawsuits on the issue.
In what appears to be an attempt to address some of these concerns, the company said in its system card for 4o image generation it had “added a refusal which triggers when a user attempts to generate an image in the style of a living artist.”
However, Studio Ghibli’s co-founder, Hayao Miyazaki, is still alive. Miyazaki has even taken issue with AI in the past, famously calling AI an “insult to life itself” in a 2016 documentary.
The free version of ChatGPT notably refuses to generate images in the style of the studio, calling it a violation of its policy guidelines.
When prompted by Fortune to create an image in the style of Studio Ghibli the chatbot said: “I wasn’t able to generate the image because the request didn’t follow content policy guidelines. If you’d like, I can create something similar with a different approach—perhaps a symbolic representation of technology and art clashing in a fantasy setting. Let me know how you’d like to proceed!”
Representatives for OpenAI did not immediately respond to a request for comment about this issue. The company told TechCrunch that while ChatGPT refuses to replicate “the style of individual living artists,” OpenAI does permit it to replicate “broader studio styles.”
