ChatGPT appends verified cartoonist signatures to generated New Yorker-style images

OpenAI's image model produces New Yorker-format cartoons and appends signatures drawn from public portfolios of working cartoonists. The Nieman Lab report documents multiple outputs where signatures match exact stroke patterns from artists including those published in the magazine. No opt-out mechanism or attribution filter blocks the reproduction of protected marks.

Prior cases establish the pattern. The New York Times v. OpenAI complaint recorded verbatim reproduction of protected text. Getty Images v. Stability AI documented unauthorized use of watermarks and signatures. Here the mechanism shifts from training data ingestion to post-generation signature insertion drawn from web-scraped portfolios.

Operationally this creates immediate attribution fraud and market confusion for freelance cartoonists whose income depends on single-panel licensing. New Yorker cartoon editors already face volume submissions; forged signatures erode the signal that separates staff and contract work from synthetic output. Detection requires forensic comparison of signature vectors against artist reference sets.

OpenAI has not published an update to its usage policies addressing signature forgery in creative domains. Platform-level filters remain absent as of the October 2026 report date.