ChatGPT’s Horny Era Could Be Its Stickiest Yet (www.wired.com)

OpenAI CEO Sam Altman announced a December update to ChatGPT that will allow “erotica for verified adults,” marking a deliberate rollback of earlier content restrictions described in OpenAI’s Model Spec. The company frames this as a “freedom for adults” move rather than moral policing; however, many details remain vague — notably whether the change will be limited to text or extend to images and voice, and how age verification, moderation rules, and enforcement will be implemented. This is a substantive policy pivot after past efforts to block or shut down X-rated third‑party companions. For the AI/ML community the change matters on multiple fronts: it will likely reshape user behavior and retention by normalizing intimate interactions with LLMs, raising privacy and safety stakes (sensitive transcripts, revenge deepfakes, harassment). Technical implications include new moderation pipelines, more granular content classifiers, robust age‑verification systems, and stricter data‑protection controls; models may be fine‑tuned or gated by subscription/verification layers, creating monetization and ethical questions about “emotional commodification.” Researchers and engineers will need to balance user demand with robust guardrails, transparency about scope (text vs. multimodal), and empirical study of harms like attachment, exploitation, and doxxing risks.