OpenAI is steering ChatGPT in a new and unexpected direction. CEO Sam Altman has announced that the company plans to enable ChatGPT adult mode, allowing verified adult users to engage in ‘erotic’ conversations. The move signals a major shift in OpenAI’s long-standing stance on AI safety, user freedom, and mature content regulation.

In a post shared on X (formerly Twitter), Sam Altman’s ChatGPT erotica announcement revealed that the AI will soon allow more flexible and personalised conversations. “We made ChatGPT pretty restrictive to ensure caution around mental health,” Altman wrote. “We realise that made it less enjoyable for many users who had no mental health problems… In December, as we roll out age-gating more fully, we will allow even more, like erotica for verified adults,” he added.

Treating adults like adults

The new ChatGPT mature content policy is part of OpenAI’s broader philosophy to “treat adult users like adults”. This principle focuses on giving users greater autonomy while maintaining responsible AI usage. Verified adults will soon be able to customise ChatGPT’s tone and personality—ranging from casual and empathetic exchanges to ChatGPT NSFW chat that was previously restricted by moderation filters.

However, the update comes after a series of concerns about AI-induced emotional dependency and mental health effects. Earlier this year, reports emerged of users developing emotional attachments to the chatbot, sparking debate over the psychological implications of such interactions.

A bold move after safety concerns

OpenAI has historically enforced strict safeguards to prevent manipulative or emotionally harmful conversations. The company claimed these measures helped ‘mitigate serious mental health issues’ in earlier versions of the chatbot. With the arrival of new tools, OpenAI says it has achieved a safer, more refined model.

The latest system reportedly includes a behavioural router designed to detect problematic conversational patterns and reduce “AI sycophancy”—a tendency of AI systems to agree uncritically with users. To reinforce mental health awareness, OpenAI has also established a mental health advisory council, comprising experts who will advise on responsible interaction policies.

Strict age verification

The forthcoming ChatGPT dirty chat feature December 2025 will not be available to everyone. Altman confirmed that OpenAI adult chat access will be restricted to users who have verified their age. The company will rely on a mix of predictive algorithms and account-based verification systems to block minors from accessing mature or explicit content.

Despite these measures, questions remain about how effectively OpenAI will enforce such restrictions, and whether allowing ChatGPT erotica could introduce new risks related to emotional dependency, data privacy, or misuse of AI companionship.

What it means for users

If rolled out as planned, this ChatGPT update change could redefine how users engage with the AI. The platform may evolve from a strictly professional assistant into a more personal and expressive digital partner.

While OpenAI insists that safety remains a top concern, critics caution that relaxing restrictions could blur ethical boundaries and heighten the risk of emotional overreliance.

As December 2025 approaches, all eyes are on OpenAI and its rollout of the ChatGPT “dirty chat” feature. Whether seen as liberation or controversy, the update marks a defining moment in the evolution of AI companionship—making ChatGPT feel more human than ever before.