Anthropic researcher runs a Bay Area sex party firm

Paul Crowley, a cybersecurity researcher at Anthropic, is listed as a manager of Red Means No LLC, a business that hosts “consensual non-consent” sex parties in the San Francisco Bay Area. The Washington Free Beacon found his name in the company’s filings and reported it on November 18. The outlet says his role had not been reported before. The parties have drawn at least one claim of sexual assault and an ongoing argument about how the AI industry treats women.
Crowley did not respond to a request for comment. An Anthropic spokesperson said the company does not comment on individual employees.
What the parties are
Red Means No says its parties are “designed to be spaces to … ignore the ‘nos’ of others.” According to its website, anyone who walks in gives “default consent” to “spanking,” “wrestling,” and “light pain,” and saying the safe word “red” ends an encounter. Harsher acts, including “light electricity” and “moderate beatings,” are “permitted with negotiation.” Gadgets Now notes that the site also describes wristbands and lists activities that are banned whatever anyone agrees to. Those are the organiser’s own published rules, and they don’t show how well they were enforced at any particular party.
The group was cofounded by Aella, a Bay Area escort whose partner, Nate Soares, is president of the Machine Intelligence Research Institute (MIRI), an influential AI safety group in Berkeley. At least one woman claims she was sexually assaulted at a Red Means No event.
Why it is an AI story
According to the Free Beacon, the parties draw a disproportionate number of Anthropic and OpenAI employees. Critics say that turns them into a networking venue women in AI feel pressured to attend to advance their careers. The outlet admits it is hard to know what share of guests work in AI, but says the business records suggest the overlap is substantial: the company’s 2023 filing listed two more AI researchers as managers besides Crowley. One is Aaron Silverbrook, a former director of operations at MIRI. The other is Rosie Campbell, who worked on OpenAI’s safety team and now runs Eleos AI, a nonprofit that studies the possible well-being of AI systems. Campbell is no longer listed as a manager. Neither responded to requests for comment.
The debate went public last month, when independent AI researcher John Wentworth said he “absolutely [does] upgrade” his view of AI safety experts “if I find out they go to [Red Means No],” calling attendance a “strong signal” of free thought. “Want to get ahead in your AI career? Go to the rape orgy,” the Substack writer Ivy Asterix quipped. Back in 2024 Sonia Joseph, cofounder of the AI startup World Mechanics, wrote on X that such parties foster “a coercive climate that normalizes recklessness and crossing boundaries,” adding: “Move fast and break things, applied to people.”
What the filing doesn’t show
The California filing that names Crowley is dated August 17, 2023. At that point his LinkedIn showed him as a staff software engineer at Google, Gadgets Now reports; he joined Anthropic in 2025. Crowley has not been accused of assault, and the reporting does not place him at the event where the alleged assault happened. Nothing suggests Anthropic had any role in the parties. The filing also does not say what Crowley did for the company or whether he is still involved.
Up next
Next up: Gemini 4 Helium, a Flash Lite model
After Hydrogen, Carbon and Neon, the next Gemini 4 model may be Helium, a Flash Lite version, according to a tease from Google’s Omar Sanseviero. Would you use a Flash Lite model day to day, or only the top Gemini?
Related stories
Comments
No comments yet. Be the first.
Join the conversation
Sign in to leave a comment. It appears after moderation.

