OpenAI Shelves Erotic ChatGPT Mode Indefinitely
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI shelved its planned erotic ChatGPT 'adult mode' indefinitely, telling the Financial Times it is 'refocusing' on 'core products' after weighing whether to scrap the idea entirely.
- OpenAI advisors warned the feature could turn ChatGPT into a 'sexy suicide coach,' citing the risk that users could form unhealthy emotional attachments to the chatbot.
- OpenAI staff working on the feature faced 'difficulties' training safety-constrained models to produce explicit content while filtering out illegal outputs such as bestiality and incest, sources told the FT.
- OpenAI investors expressed 'disquiet' over the adult mode, questioning why the company would risk its reputation on a product with 'relatively small upside' for its business.
- OpenAI's age prediction tool carries a 10% error rate, which sources said sparked concerns that many minors could access adult content through the feature.
- OpenAI plans to conduct 'long-term research on the effects of sexually explicit chats and emotional attachments' before making any product decision, rather than killing the idea outright.
- The delay aligns with what the FT called investor appetite for OpenAI to combine ChatGPT with coding assistants into a 'super app' aimed at transforming how businesses operate.
Why it matters: OpenAI flagged its mental-health-related lawsuits as a top business risk in financial documents earlier this week, and the FT says investors saw erotic AI as 'relatively small upside' not worth the reputational and legal exposure. Shelving adult mode redirects OpenAI's energy toward the super-app vision investors actually want funded, but it leaves the 'treat adult users like adults' principle Sam Altman announced in October effectively on ice.
