OpenAI rolled out a new image generator powered by its GPT-4o model—but pulled it from the free version of ChatGPT just one day later.
Why? The feature took off fast, with users creating a flood of images in the style of Studio Ghibli, the famous Japanese animation studio.
The tool was meant to show off GPT-4o’s new abilities. It creates images using a left-to-right, top-to-bottom method—different from how older models work. This approach helps the AI generate more realistic pictures and sharper text inside the images, fixing one of the biggest issues users had with past versions.
OpenAI also added human feedback to help the model improve right after launch.
But things moved quickly. As soon as the update went live, people started sharing AI-generated Ghibli-style versions of everything—The Godfather, Star Wars, viral memes like “distracted boyfriend,” and more.
Even OpenAI CEO Sam Altman joined in, switching his X (formerly Twitter) profile picture to a Ghibli-style portrait of himself.
Later that same day, Altman announced that OpenAI was temporarily pausing the image generation tool for free users. He didn’t say when it would come back.
The feature is still available for paid users on ChatGPT Plus, Pro, and Team plans.
Why OpenAI Paused the Feature
The quick move to pull the feature suggests OpenAI wasn’t expecting such a huge response—especially from people generating images that closely mimicked a well-known visual style.
While styles themselves aren’t usually protected under copyright law, using them in this way raises questions about what’s fair, legal, or ethical when AI is involved.
OpenAI has faced similar concerns before, and this latest incident shows how tricky it is to balance powerful tools with responsible use.
For now, the company seems to be stepping back to figure out how to move forward without crossing any legal or creative lines.
