ChatGPT is adding real cartoonists' signatures to fake New Yorker cartoons
Points and comments are a snapshot, not live.
ChatGPT generates New Yorker-style cartoons with real cartoonists' signatures without permission.
AI-generated New Yorker cartoons featuring signatures of real cartoonists like Brendan Loper and Emily Flake have circulated on social media. Loper called it a violation of his personhood. OpenAI said it appreciates bug reports but hasn't stopped the behavior. Condé Nast's deal with OpenAI doesn't cover cartoons; cartoonists retain copyright. Jason Adam Katzenstein compared it to false impersonation. Legal experts say copyright claims are weak because styles aren't specific works, but right of publicity cases may apply.
What commenters are saying
Commenters split on whether this is a bug or forgery. Some argue it's a trademark or copyright issue, while others say the model statistically reproduces signatures from training data without intent. A top commenter notes attribution is different from forgery: humans know not to sign others' work, but LLMs just generate likely pixels. Another camp suggests the real problem is stolen training data, not the signature output itself.