ChatGPT caricatures are taking over social media—but at what cost?

Have you seen larger-than-life depictions of your friends lately? They might have been sucked into the latest social media trend: creating AI-generated caricatures.

The trend itself is simple. Users input a common prompt—“Create a caricature of me and my job based on everything you know about me”—then upload a photo of themselves, and voilà! ChatGPT (or any AI-image platform) spits out an over-the-top cartoon-style image that includes features from their job, their favorite activities, and anything else it’s learned about them.

This ability is predicated on a robust ChatGPT (or other AI) chat history. Those who don’t have a close, personal relationship with the AI might need to give additional information to get a more accurate depiction. But, notably, that makes for yet another potential AI privacy concern.

It’s not the first AI image trend. Other social media challenges have had users posting themselves as AI-generated cartoons, Renaissance paintings, or fantasy characters.

AI’s image capabilities have gone in a few different directions. Some of them—like with this trend or with the meme-ification of Sora—are seemingly harmless fun. However, Sora has started to see issues with bad-faith individuals being able to create AI deepfakes. (See also: Grok porn.)

Meanwhile, even as the trend continues to rise, more than 13,000 ChatGPT users reported issues on Thursday, according to outage tracking website Downdetector.com.