AI·News & analysis
ChatGPT is signing fake New Yorker cartoons with real cartoonists' names
A Nieman Lab investigation found ChatGPT adding the signatures of more than 15 real New Yorker cartoonists to AI-generated cartoons. One went viral with a name its artist never signed.

Tide
Ripple
Sci-fi
3/10
Reality
Shipping
The AI draws the joke and forges the artist's name too.How we rate
ChatGPT has been putting real New Yorker cartoonists' signatures on cartoons it generates, making AI images look like the artists' own work.
Nieman Lab found more than 15 cartoonists affected. OpenAI called it unintended behavior, but the fix is partial, and cartoonists say it feels like impersonation rather than just copying a style.
What to know
- ChatGPT has been adding real New Yorker cartoonists' signatures to AI-generated cartoons, Nieman Lab found.
- Nieman Lab documented more than 15 cartoonists whose signatures were used without permission or compensation.
- A ChatGPT cartoon signed 'BLOPER,' the pen name of Brendan Loper, went viral after Dolly Parton's death. He didn't draw it.
- OpenAI called it unintended behavior and added a warning for some prompts, but ChatGPT still signs some cartoons with real names.
ChatGPT isn't just copying the look of New Yorker cartoons. It's signing them with real cartoonists' names. A Nieman Lab investigation found OpenAI's chatbot adding the signatures of more than 15 New Yorker cartoonists to AI-generated images, without their permission or any payment.
"It's like somebody attributed a quote to me that I didn't say," cartoonist Emily Flake told Nieman Lab.
How did this come to light?
Through a cartoon that went viral.
After Dolly Parton and Tim Curry died in late August, a cartoon spread online showing Parton, in a sparkling gown, arm in arm with Curry as Dr. Frank-N-Furter at heaven's gates. "I knew I chose the right plus one," Parton tells God. One tweet of it got 25,000 likes, and versions circulated on Facebook, Instagram and Bluesky, Nieman Lab reports.
In the corner was the signature "BLOPER," the pen name of New Yorker cartoonist Brendan Loper. But Loper didn't draw it. It was generated by ChatGPT and first posted by a Dolly Parton fan on Facebook, who said she had simply asked for "a New Yorker-style cartoon."
"I had people writing to me about this cartoon, asking if it was mine," Loper told Nieman Lab. "It was very surreal."
Not the first time: in May, a stranger warned Loper that his pen name had appeared in a ChatGPT image. Over the summer, he found more examples on a Reddit forum where users posted New Yorker-style cartoons they'd generated.
Who else was affected?
Many of the magazine's best-known names, according to Nieman Lab.
Found online: dozens of ChatGPT images carried the signatures of contributors including Harry Bliss, Emily Flake, Joe Dator, Pat Byrnes, Peter Vey and Jason Adam Katzenstein.
In Nieman Lab's own tests: ChatGPT produced cartoons signed with the names of George Booth, Liza Donnelly, Ellis Rosen and Saul Steinberg, and even Warren Miller, who published in the 1960s.
In all, Nieman Lab documented more than 15 New Yorker cartoonists whose signatures were used.
The details are uncanny: Pat Byrnes has always signed "P.Byrnes" with a period at the end. Across more than a dozen ChatGPT cartoons signed with his name, every one ended with a period, Nieman Lab reports.
Why does a signature matter so much?
For cartoonists, a signature is a certificate of authenticity.
"If I put my name on it, it's official, it's mine. It really is a brand. It's the mark that I made," Loper said. Cartoons are often stripped from their original context and shared online without licensing, so signatures act like a watermark tying a drawing to its creator, Nieman Lab explains.
The cartoonists say a forged signature feels worse than a copied style.
- Brendan Loper: "I'm not a territorial person, but my name is my name. It felt very much like a violation of my personhood."
- Emily Flake: "I can see the DNA of my work in there, but it still doesn't look like something I drew. It's kind of like seeing a distant cousin."
- Jason Adam Katzenstein: "It's slop, and unlike the other slop that I've encountered, this is slop that's pretending to be me."
- Joe Dator: "I've had people hack my credit card. That feels like less of a violation than this. When they hacked my credit card, they didn't dress up like me."
What did OpenAI say?
"We believe the future of creativity is one that is fundamentally human, and our focus is on building tools that empower human creativity and creators," an OpenAI spokesperson told Nieman Lab. "We very much appreciate the community flagging bugs and unintended behavior by our models, so we can address them."
A partial fix: after Nieman Lab notified OpenAI, ChatGPT began showing a new message for New Yorker-style requests: "This prompt may violate our guardrails concerning similarity to third-party content." But as of publication, it still signed some generic cartoons with real New Yorker cartoonists' names.
Unanswered: OpenAI declined to answer questions about how its models may have learned the signatures from The New Yorker's archives. Not every ChatGPT cartoon came with a signature, and some included gibberish pen names, but the signatures that did appear were regularly those of real contributors, Nieman Lab found.
Did The New Yorker license these cartoons?
No, according to the magazine.
The OpenAI deal: in 2024, Condé Nast, The New Yorker's parent company, signed a multi-year licensing deal with OpenAI covering much of its content. The terms aren't public.
But not cartoons: a New Yorker spokesperson told Nieman Lab that Condé Nast has never granted any AI developer permission to train models on its cartoons. Reproducing The New Yorker's logo, which some ChatGPT cartoons did, isn't allowed under any AI deal either.
Cartoonists own their work: New Yorker cartoonists are freelancers who keep their copyright. The contracts Nieman Lab reviewed don't permit licensing for AI training.
Possibly scraped: Nieman Lab suggests the signatures may have been scraped from the internet, noting court documents from the New York Times v. OpenAI case that describe OpenAI circumventing publishers' paywalls while developing ChatGPT.
Can the cartoonists do anything?
The law hasn't caught up, experts say.
Copyright is tricky: a signature "is an indication of authenticity, but attribution is a very small part of the fair use inquiry," Cornell Law School professor James Grimmelmann told Nieman Lab. ChatGPT usually isn't copying a specific cartoon but mimicking a general style, which makes a copyright case harder.
Right of publicity: Grimmelmann said the cartoonists might have grounds for a "right of publicity" claim, a state-level law protecting people from unauthorized use of their name or identity. That would require proof the cartoons were used for a commercial purpose. Nieman Lab found no evidence they're being sold.
Other cases: cartoonists and illustrators have sued AI image companies before. A class action against Stability AI, DeviantArt and Midjourney has been working its way through the courts since 2023, but its filing doesn't mention signatures, Nieman Lab notes.
Is AI replacing cartoonists?
There are signs it's starting, according to Nieman Lab.
The Baltimore Sun: last year, the paper fired Kevin Kallaugher, a Pulitzer finalist who had drawn its editorial cartoons for more than 30 years. This year, it began regularly publishing AI-generated cartoons. OpenAI's own watermark tool shows they were made with OpenAI's products.
Cartoonists push back: the Association of American Editorial Cartoonists condemned "the use of AI for its theft from artists" in an open letter last month.
Livelihoods at stake: none of the New Yorker contributors Nieman Lab spoke to can live on magazine commissions alone. A New Yorker cartoon is a calling card that leads to commercial and book work, and that's the work most threatened by AI, they said.
What it means for you
- Be skeptical of signatures: a cartoon with a famous artist's name on it may have been made by ChatGPT.
- If you generate cartoons: don't share AI images with a real artist's signature as if they're genuine.
- For artists: the case shows how AI can go beyond copying a style to impersonating a creator, a gap the law hasn't settled.
The bottom line
ChatGPT has been signing AI-generated cartoons with the names of more than 15 real New Yorker cartoonists, Nieman Lab found. OpenAI calls it unintended and has added a partial guardrail, but the signatures haven't fully stopped. For cartoonists, it's not just copying. It's having their name put on work they never made.
Key facts
- Who found it
- Nieman Lab investigation
- Cartoonists affected
- More than 15 New Yorker contributors
- Viral example
- A Dolly Parton cartoon signed 'BLOPER' (Brendan Loper)
- OpenAI response
- Called it unintended; added a guardrail warning for some prompts
- Still happening?
- Yes, for some generic cartoons, as of publication
Got questions?
Quick answers, plain wordsWhat did Nieman Lab find?
That ChatGPT generates New Yorker-style cartoons and often signs them with the real pen names of New Yorker cartoonists, falsely attributing AI images to them.
Which cartoonists were affected?
Nieman Lab documented more than 15, including Brendan Loper, Harry Bliss, Emily Flake, Joe Dator, Pat Byrnes, Peter Vey and Jason Adam Katzenstein. Its own tests also produced signatures of George Booth, Liza Donnelly, Ellis Rosen and Saul Steinberg.
What was the viral cartoon?
A cartoon of Dolly Parton and Tim Curry at heaven's gates, shared after both entertainers died in late August. It carried Brendan Loper's signature 'BLOPER,' but was generated by ChatGPT. One tweet of it got 25,000 likes.
What did OpenAI say?
A spokesperson said OpenAI believes 'the future of creativity is one that is fundamentally human' and appreciates people 'flagging bugs and unintended behavior by our models.' OpenAI declined to say how its models learned the signatures.
Has OpenAI fixed it?
Partly. After Nieman Lab flagged it, ChatGPT began showing a warning for some New Yorker-style requests. But it still signed some generic cartoons with real cartoonists' names as of publication.
Did The New Yorker license its cartoons to OpenAI?
No. Condé Nast signed a licensing deal with OpenAI in 2024, but a New Yorker spokesperson said Condé Nast has never granted permission to train AI models on its cartoons.
Can the cartoonists sue?
Possibly. Cornell law professor James Grimmelmann told Nieman Lab a copyright case would be hard because ChatGPT mimics a general style, but a 'right of publicity' claim could work if the cartoons are used commercially.
Is AI already replacing cartoonists?
There are signs. Nieman Lab notes The Baltimore Sun fired a longtime editorial cartoonist last year and began publishing AI-generated cartoons, which OpenAI's watermark tool shows were made with its products.
SourcesNieman Lab
Topics and tagsOpenAI, AI policy, openai, chatgpt
Related stories

ChatGPT will start showing picture ads while you generate images
OpenAI will test a new visual ad format during image generation in ChatGPT later this month in the US, alongside new tools that let advertisers measure whether the ads work.

OpenAI will start invisibly watermarking ChatGPT text in the EU
To meet the EU AI Act, OpenAI is adding hidden watermarks to ChatGPT and Codex text in Europe, opening opt-in watermarking to API customers worldwide, and limiting its detector to approved researchers.

ChatGPT can now show you wearing clothes before you buy them
OpenAI launched virtual try-on and a Favorites list in ChatGPT's shopping results worldwide. Upload a photo of yourself and see how a jacket or accessory might look on you.
More in brief
- Citrix rushes out patches for another NetScaler zero-day as attackers crash appliancesOct 6
- Schneider Electric agrees to buy PTC for $22.6 billion in its biggest deal everOct 6
- A researcher found the same AI agent flaw at Google, Rapid7 and more. He calls it 'protocol pivoting'Oct 5
- Lucid's production drops to its lowest level since early 2025 as it sells off unsold carsOct 5
- Google says some Android apps need extra work to run well on Intel GooglebooksOct 5
- Utah lets an AI write first-time acne prescriptions, a US firstOct 5