OpenAI is taking a surprising turn. CEO Sam Altman announced that ChatGPT will soon allow erotic or explicit conversations, but only for verified adult users. The decision marks a major shift in OpenAI’s approach to AI safety and user freedom.
In a post shared on X (formerly Twitter), Altman said the company plans to relax some of ChatGPT’s long-standing safety restrictions, giving users more control over the tone and nature of their chats. “We made ChatGPT pretty restrictive to make sure we were being careful with mental health issues,” Altman wrote. “We realize this made it less useful or enjoyable for many users. In December, as we roll out age-gating more fully, we will allow even more, like erotica for verified adults.”
Treating adults like adults
The update is part of OpenAI’s broader “treat adult users like adults” principle, which aims to balance responsible AI usage with user autonomy. The company says verified adults will soon be able to customise ChatGPT’s behaviour, from being more casual and empathetic to engaging in adult-themed chats that were previously blocked by strict moderation filters.
This move, however, comes after months of scrutiny over ChatGPT’s impact on mental health and emotional dependence. Several incidents earlier this year raised alarms when users reported becoming emotionally attached to the chatbot or being led into delusional scenarios by the AI’s responses.
A controversial shift after safety concerns
OpenAI had previously rolled out multiple safeguards to prevent ChatGPT from engaging in manipulative or emotionally risky exchanges. The company said these measures helped “mitigate serious mental health issues” seen in earlier versions of the model.
Now, with the rollout of GPT-5, OpenAI believes it has achieved a safer, more stable AI. The latest model reportedly includes a behavioural router that detects concerning patterns and reduces “AI sycophancy”, the tendency of chatbots to uncritically agree with users.
To further strengthen oversight, OpenAI has also formed a mental health advisory council, made up of experts who will guide policy on well-being and AI interactions.
Age verification will be key
The upcoming “adult mode” will not be universally available. Altman confirmed it will only unlock for users who have verified their age, and the feature will remain off by default. The company plans to use a mix of age prediction systems and account-based verification to ensure that minors cannot access explicit content.
Still, questions remain about how effectively OpenAI will enforce these restrictions, and whether introducing erotic chat features could create new challenges around AI dependency, emotional manipulation, and privacy.
What this means for ChatGPT users
If implemented as described, this change could dramatically alter how users interact with ChatGPT, shifting it from a strictly professional and educational tool to a more personal and expressive digital companion.
While OpenAI insists that safety remains its top priority, critics warn that loosening restrictions may blur the lines between ethical AI use and risky emotional engagement.
For now, users will have to wait until December, when OpenAI begins rolling out the age-gated adult chat feature. One thing’s certain, the next version of ChatGPT is about to become far more human, for better or worse.
