OpenAI's ChatGPT signs fake cartoons with real artists' names

Okay so on October 6, 2026, Nieman Lab reported that ChatGPT has been putting real New Yorker cartoonists' signatures on AI-made, New Yorker style cartoons they never drew.

Whose names did ChatGPT use?

ChatGPT used the signatures of more than 15 New Yorker cartoonists, including Harry Bliss, Emily Flake, Liza Donnelly and Brendan Loper. Loper found his signature, "BLOPER," on a cartoon he never drew. Loper first heard about it in May through an Instagram message, and more examples popped up on Reddit over the summer. He called it "a violation of my personhood." Another longtime New Yorker cartoonist, Jason Adam Katzenstein, described the output as "an uncanny simulacrum of an impression of a joke."

Didn't The New Yorker license its work to OpenAI?

Condé Nast, The New Yorker's parent company, signed a licensing deal with OpenAI in 2024, but that deal excluded cartoons. So how ChatGPT learned these signatures is still unexplained. It could be older scraping, other data or the open web. Nobody has confirmed which.

What did OpenAI do about ChatGPT's fake signatures?

When the issue was raised, ChatGPT started answering that kind of prompt with a guardrails warning saying it "may violate our guardrails concerning similarity to third-party content." There's still no public explanation of where the signatures came from.

Bro. I make content for a living. Imagine your name on something you never made.

Why do fake AI signatures matter?

Fake AI signatures matter because a signature tells you who made something. Real talk: copying a style is one thing. Putting a real person's name on work they never made is another.

So don't trust a signature. Just check who really made it. Like, now.