ChatGPT signatures make fictional cartoons look authorial
The ChatGPT signatures controversy centers on a type of single-panel cartoon generated in The New Yorker's style. The images imitate the magazine's black-and-white line work and sardonic captions, while adding the pen names of real cartoonists. Brendan Loper discovered that a widely shared image bearing his signature, “BLOPER,” was not his work, yet readers contacted him as if it were.
A signature is an identity and provenance marker
The report documented signatures associated with more than 15 New Yorker cartoonists, including Emily Flake, Joe Dator, and Pat Byrnes. The artists distinguish style imitation from signature reproduction. A similar style may raise a copyright question, but a signature directly suggests that the work was created or endorsed by that person. For illustrators who rely on signatures to establish authorship amid context-free online reposting, the result can mislead audiences and damage a personal brand.
OpenAI added a warning, but the source and responsibility remain unclear
OpenAI said the behavior was unintended and should be addressed. After being notified by the reporter, ChatGPT began warning that requests for New Yorker-style cartoons might violate safeguards concerning similarity to third-party content. At publication, however, the system still attached real cartoonists' names to some generic images, showing that the protection was inconsistent. The source of the signatures also remains unresolved: the Condé Nast deal with OpenAI is not public, The New Yorker says it never authorized training on its cartoons, and the available reporting does not establish a specific ingestion path.
The central issue is not only whether AI can draw similarly, but whether generated content presents a nonexistent author relationship as fact. U.S. intellectual-property law has not fully caught up with signature impersonation in generative images, leaving the boundaries among copyright, false attribution, and platform safeguards unsettled. For publishers, platforms, and model developers, limiting style imitation is only one step. They also need to prevent real names or pen names from being attached to unauthorized work and provide affected creators with a verifiable way to challenge such outputs.