AI has come for comics on stage and in The New Yorker

The first time George Brett Olson saw a woman on a screen mimic his punchline, he thought it was a joke. The womanan AI avatar named Tiffany Sloanhad the same exaggerated gestures he used, the same cadence, even the same bright orange hoodie he wore onstage. When he watched the clip on a social‑media feed, the realization hit him like a punch: someone had taken his set, fed it into a model, and turned it into a new character that was now earning money on an adult‑content platform. The clip that made him aware of the theft was shared by writer Nate Bowling in his Takes & Typos newsletter, and it quickly spread through the comedy community.

Stand‑up, the line between inspiration and theft is already thin. A joke shared on a podcast or a tweet can become a meme in minutes, and the original creator often loses control. With generative AI, that loss of ownership is amplified. Large language models are built on vast amounts of data, and the more content they ingest, the more convincing their output becomes. Olson’s experience is not an isolated case; other comedians have reported similar incidents, but the visibility of his story has brought the issue into sharper focus.

Olson’s video shows Sloan’s avatar mimicking his stage presence with uncanny precision. The model’s creators used his jokes as motion capture, treating him like a performer and her like a digital double. The result is a character that not only steals his material but also monetizes it through an OnlyFans page. The irony is that the content that sparked the theft was posted on a platform that, according to its own policies, does not prohibit such use. The platform’s relaxed moderation allows the AI to harvest and remix content without the original creator’s consent.

While stand‑up comics are grappling with AI‑generated avatars, the world of editorial cartoons is facing a new threat. A report from Nieman Labs revealed that ChatGPT has been producing New Yorker‑style cartoons that bear the signatures of real artistsBrendan Loper, Harry Bliss, Emily Flake, Pat Byrnes, George Booth, and Liza Donnelly, among otherswithout their knowledge. For cartoonists, a signature is more than a name; it is a brand that signals authenticity and carries monetary value. When an AI produces a cartoon and appends a real artist’s signature, it not only misrepresents the creator but also dilutes the market for genuine work.

The situation is complicated by a licensing deal OpenAI signed with Condé Nast, the parent company of The New Yorker, in 2024. The agreement explicitly excluded cartoons from training data, yet the AI still generates cartoons that mimic the magazine’s style and, disturbingly, use real signatures. The result is a product that looks like a legitimate New Yorker illustration but is, in fact, a synthetic creation.

Both casesOlson’s stand‑up avatar and the New Yorker cartoonsillustrate a broader economic pattern. The creators of the AI content are reaping significant financial rewards. Sloan’s OnlyFans page, for example, earns revenue from viewers who are drawn in by the familiar jokes and movements. The cartoon generators, meanwhile, can produce thousands of images at a fraction of the cost of hiring a human artist. The creators of these AI tools are amassing a following and directing traffic to monetized platforms, all while sidestepping the legal and ethical boundaries that protect original work.

This economic incentive is a driving force behind the proliferation of AI‑generated content. The more profitable the model, the more it will be trained on existing works, and the more it will replicate them. The cycle becomes self‑reinforcing: the more the AI is used, the more it learns, and the more it can produce convincing copies of original content.

The incidents involving Olson and the New Yorker cartoonists underscore the urgent need for clearer guidelines around AI training data and the use of copyrighted material. While platforms may claim that user‑generated content is public domain for training purposes, the reality is that creators still hold rights to their work. The current lack of enforcement allows AI developers to bypass these rights, creating a market for stolen content that is difficult to trace and difficult to stop.

As a writer who has spent years watching comedians and cartoonists hone their craft, I see the frustration in their eyes. They have spent years building a brand, a voice, and a loyal audience. To have that hard‑earned identity turned into a commodity for an anonymous AI developer is not just a violation of intellectual property; it is a blow to the creative community’s sense of agency.

The next few months will likely see more legal challenges and policy debates as creators demand protection from AI‑generated theft. If the industry moves quickly to establish robust licensing frameworks and enforce existing copyright laws, it could safeguard the livelihoods of comedians and cartoonists alike. Until then, the line between inspiration and appropriation will continue to blur, and the creative community will have to navigate a landscape where the very tools that can amplify their reach also threaten to undermine it.