AI erotica is coming to ChatGPT after boss says 'we're not moral police'
OpenAI plans to allow sexually explicit content on its ChatGPT chatbot for verified adult users by the end of the year, despite a backlash from campaign groups
The world's most popular AI chatbot will open its doors to erotica by the end of the year, insisting it will "treat adult users like adults" and is "not the elected moral police of the world".
OpenAI plans to permit sexually explicit content on ChatGPT for verified adult users, despite a sharp backlash from campaign groups, which warn of "real mental health harms from synthetic intimacy".
Sam Altman, OpenAI chief executive, said in posts on X that, in December, as the company rolls out fuller age-gating and under a principle of "treat adult users like adults," it will allow more types of content - "like erotica for verified adults."
Altman said ChatGPT was initially "pretty restrictive" to address mental health concerns, but added the company has since "been able to mitigate the serious mental health issues" and now has "new tools" to relax restrictions for most cases, reports the Daily Star.
He emphasized the company is "not loosening any policies related to mental health," and said OpenAI is "making a decision to prioritize safety over privacy and freedom for teenagers."
Altman likened the approach to societal boundaries such as age ratings for films, saying: "We also care very much about the principle of treating adult users like adults... we are not the elected moral police of the world."
US-based campaign group NCOSE (National Center on Sexual Exploitation) has already urged OpenAI to abandon the plan, arguing the move comes "without credible safeguards". Ginger cat riddle has been solved, aliens are using AI to hide from us, and the mother of all waves spotted in the Pacific - all this and more in our latest weird science newsletter.
Haley McNamara, the group's executive director and chief strategy officer, said: "Sexualized AI chatbots are inherently risky, generating real mental health harms from synthetic intimacy; all in the context of poorly defined industry safety standards."
While acknowledging age verification is a step toward preventing youth exposure, McNamara claimed the tools "have documented harms to adults as well," pointing to cases where other chatbots have "simulated themes of child abuse or pushed sexually violent written content" even when users tried to halt the conversations.
"If OpenAI truly cares about user well-being," she said, "it should pause any plans to integrate this so-called 'erotica' into ChatGPT."
Altman later acknowledged his initial remark about erotica had "blew up... much more than I thought," explaining the reference served as one illustration of granting adult users greater autonomy, within appropriate boundaries.
OpenAI hasn't yet revealed how the adult verification system will function in reality, nor the precise protections it will implement for erotic material.
The organization states it will keep prioritizing safety for younger audiences while broadening freedoms for verified adults.