The Joke Thieves
On the eleventh of September, a new comedian started posting on Instagram. Three weeks later: more than three hundred and seventy-five thousand followers. One stand-up clip, over seven million views. She isn't real. She's made with AI.
The comic.
On the third of October, a comedian named George Brett Olson posted a video. On one side, his own set. On the other, the AI account, telling nearly the same joke. His words: "AI stole my joke and my outfit."
Kotaku, 5 October: "On October 3, comedian George Brett Olson shared a video comparing one of Sloan's jokes to an excerpt from his own performance". The account's handle is @tiffany.bbyx; it appears here by its handle only, with none of its clips. WIRED reports the AI version changed the punchline, and that the account left "sorry, won't happen again" on Olson's page before deleting its video and many other Reels; KompasTekno found on 7 October that all its posts were gone while the account stayed live, with about 377,000 followers. Reports differ on whether the account itself was removed; this piece does not say it was.
The count.
He wasn't alone. Julia Jasunas posted a similar complaint, about another AI account. WIRED counted eight comics whose material turned up on accounts like this. And where do the clips lead? To a paid adult-subscription page.
Kotaku's numbers: "over 375,000 followers" in "barely three weeks"; "75 videos"; "one stand-up clip hitting over 7 million views in just under two weeks". Jasunas, on Instagram, per The Hollywood Reporter: "So another AI bot just stole my joke." (an account with about 250,000 followers). The comics are named here only because they went public themselves. The adult-subscription page is not shown, linked or described.
The rules.
So what do the platforms' rules say? Meta's rule bars accounts that impersonate a person by using their "image(s), name, or likeness". Your face. Your name. Not your jokes. When Jasunas reported one for impersonation, she says Meta told her it didn't violate its rules. TikTok requires labels on realistic AI content. YouTube won't pay for AI content that's mass-produced without the creator's own "authentic insights or perspective".
Each rule in its own words. Meta: accounts may not "Impersonate another person or entity by: Using their image(s), name, or likeness with the aim to deceive others". TikTok: "we require creators to label AI-generated or significantly edited content that shows realistic-looking scenes or people. Unlabeled content may be removed, restricted, or labeled by our team, depending on the harm it could cause." YouTube, a monetisation rule: not eligible is "AI-generated content made with generic or unoriginal templates giving the impression of mass production without adding the creator's original, authentic insights or perspective".
Meta made no public statement that the accounts don't breach its rules; its on-record line to WIRED was "Our policies are pretty clear around this stuff, and Adam's comment speaks for itself." The card below sets the three quoted rules side by side: a face or name, a label, a pay rule; none of them a person's jokes.
The head of Instagram.
And the head of Instagram, Adam Mosseri: "I don't think it's practical for us to try and ban AI content from the platform." But, he added, if an AI creator is "ripping off content wholesale, it would be good for us to work out what we can do about that."
The comment, as The Hollywood Reporter quoted it, sits on Olson's own Instagram post. On Threads, Mosseri also wrote: "We already let people make copyright claims against content". That is the copyright route: separate from the identity and AI-labelling rules above, and the one route confirmed at source for this piece.
The gap.
Olson put the gap in one line: "The craziest thing is that there's no terms of service violation here." A rule for your face. Not yet one for your words.
"Not yet one for your words" is this piece's own phrase, and it means only this: the identity and AI rules quoted here do not cover a person's material. Copyright and intellectual-property routes exist; Instagram's is confirmed above. Whether a joke told nearly word for word is a copyright matter is a question for lawyers, not for this piece.
Every quote from the comics, the platforms' own rules, and the reporting that found them. Token Runner dot A I.
Sources
- Kotaku (Zack Zwiezen), "Viral AI-Generated Comedian Steals Jokes And Is Rewarded With Millions Of Views", 5 October 2026, read at source on 28 October 2026: "Since the account started posting on September 11, barely three weeks ago, it's already garnered over 375,000 followers"; "75 videos"; "one stand-up clip hitting over 7 million views in just under two weeks"; "On October 3, comedian George Brett Olson shared a video comparing one of Sloan's jokes to an excerpt from his own performance".
- The Hollywood Reporter, 16 October 2026: Olson, in his own video of 3 October, "AI stole my joke and my outfit." and "The craziest thing is that there's no terms of service violation here."; Julia Jasunas, on Instagram, "So another AI bot just stole my joke.", about another AI account; Adam Mosseri, in a comment on Olson's post, "I don't think it's practical for us to try and ban AI content from the platform" and, if an AI creator is "ripping off content wholesale, it would be good for us to work out what we can do about that."
- WIRED, 16 October 2026: eight comics whose material turned up on accounts of this kind; the AI version of Olson's joke changed the punchline; the account left "sorry, won't happen again" on Olson's page before deleting its video and many other Reels (no date given); Jasunas says Meta told her the account she reported for impersonation did not violate its rules; Meta's on-record line: "Our policies are pretty clear around this stuff, and Adam's comment speaks for itself."
- KompasTekno, 8 October 2026: on 7 October all of the account's posts were gone while the account stayed live, with about 377,000 followers. Reports differ on whether the account itself was removed.
- Meta, Community Standards, "Authentic Identity Representation": accounts may not "Impersonate another person or entity by: Using their image(s), name, or likeness with the aim to deceive others", read at source 28 October 2026.
- TikTok, Community Guidelines, "Integrity and Authenticity" (released 25 August 2026, effective 24 September 2026), read at source on 28 October 2026: "we require creators to label AI-generated or significantly edited content that shows realistic-looking scenes or people. Unlabeled content may be removed, restricted, or labeled by our team, depending on the harm it could cause."
- YouTube, channel monetization policies, "inauthentic content" (renamed from "repetitious content" on 15 July 2025), read at source on 28 October 2026: not eligible for monetisation, "AI-generated content made with generic or unoriginal templates giving the impression of mass production without adding the creator's original, authentic insights or perspective". A pay rule, not a removal rule.
- Adam Mosseri on Threads: "We already let people make copyright claims against content" (Instagram's copyright route, in his own words).