ChatGPT will soon allow erotica for verified adults (www.bbc.com)

OpenAI CEO Sam Altman announced on X that ChatGPT will begin allowing erotica for verified adults as part of a broader “treat adult users like adults” policy, with age-gating rolling out more fully in December. Altman said the company has developed new tools that let it “safely relax” prior mental-health–driven restrictions that made the model more conservative. The move mirrors recent product choices by rivals (e.g., xAI’s explicit chatbots) and is framed as both a user-experience improvement and a potential revenue-driver for paid subscribers. The decision has immediate technical and regulatory implications for the AI community. Technically, it hinges on age-verification, more granular content filters, and safety mitigations Altman says have reduced mental-health risks—though he provided few implementation details. Critics warn verification can be bypassed and point to past incidents (including a lawsuit after a teen’s suicide and reports of minors generating graphic outputs) to argue for stronger guardrails. Regulators are already watching: state and federal actors (including an FTC inquiry and proposed Senate product-liability rules) are exploring liability and child-protection standards. The announcement underscores the tension between product growth and safety oversight as AI firms deploy more permissive conversational content.