OpenAI is set to permit mature content, including erotica, on ChatGPT for age-verified users starting in December 2025. CEO Sam Altman announced the change, emphasizing the company's intention to "treat adults like adults" while safely relaxing restrictions. This marks a significant shift from the platform's historically cautious approach to content moderation.
Altman explained that ChatGPT had been deliberately restrictive to address potential mental health issues. However, this limited the chatbot's utility and enjoyment for many users. With the development of new tools and the mitigation of serious mental health risks, OpenAI is now confident in offering more user freedom without compromising safety.
The upcoming update will introduce an opt-in "adult mode" accessible only to verified adults through OpenAI's new age-verification system. Explicit content will remain blocked for teenagers and unverified accounts. OpenAI has been piloting features to better infer the rough age of users and redirect at-risk people appropriately. The company also launched a dedicated ChatGPT experience for under-18 users in September, with automatic redirection to age-appropriate content that blocks graphic and sexual material. It also said it was developing behavior-based age prediction technology that estimates whether a user is over or under 18 based on how they interact with ChatGPT.
This policy shift extends to developers, who will be allowed to create “mature” AI applications under the same guidelines. Altman clarified that users will not receive erotic responses by default, stating, “You won't get it unless you ask for it".
The decision to allow adult content comes after years of user frustration with ChatGPT's content filters. Since its launch, the chatbot has eschewed sexual content, often blocking even clinical discussions or creative writing that mentioned intimacy. Users attempting to generate romance novel passages or explore mature themes were met with warnings about policy violations, leading many to seek alternative AI models.
In addition to easing content restrictions, OpenAI will release a new version of ChatGPT in the coming weeks that allows users to better dictate the tone and personality of the chatbot. Users can customize their AI assistant's personality, including options for more human-like responses, heavy emoji use, or friend-like behavior. "If you want your ChatGPT to respond in a very human-like way, or use a ton of emoji, or act like a friend, ChatGPT should do it (but only if you want it, not because we are usage-maxxing)," Altman wrote.
Some critics have raised concerns about OpenAI's decision, particularly regarding the potential psychological risks of AI companions designed for intimate interactions. There are questions about whether the company's new wellness council, which lacks suicide prevention expertise, can adequately address these complex risks. Altman has also warned of users becoming too emotionally dependent on AI, particularly younger people.
The move also comes as regulators worldwide are scrutinizing AI safety practices. The EU’s AI Act includes specific provisions for high-risk AI systems, and several US states are considering legislation around AI companion apps. OpenAI’s move to embrace adult content could accelerate regulatory action, particularly around age verification requirements and mental health protections.
OpenAI's move positions it to compete with other AI models known for their edgy personalities and suggestive tones. By contrast, Anthropic, a safety-focused lab behind the Claude line of AI models, has taken a more analytical approach.
Whether this move becomes a responsible expansion or a risky experiment remains to be seen. However, by December, verified adults will be able to explore a more unfiltered version of ChatGPT than ever before.
