Key Takeaways
- OpenAI will launch an adult-content capable ChatGPT version in December 2025
- Access requires strict age verification for users 18+
- New AI personality customization features coming in weeks
- Strict bans remain on illegal content and sexual violence
OpenAI CEO Sam Altman has announced plans to launch a version of ChatGPT that can handle adult-oriented content, including erotica, marking a major policy shift for the AI company. The rollout is expected by December 2025 and will require users to complete identity verification confirming they are of legal age.
From Conservative Stance to Adult Content
This move represents a significant departure from OpenAI’s historically conservative approach. The company had previously maintained stringent content limitations to protect users’ mental well-being, which many found made the chatbot overly reserved and inflexible.
According to Altman, new tooling now provides enhanced mental health safeguards, allowing OpenAI to ease previous restrictions while promising users greater freedom and safety in their interactions.
No “Sex-Bots” Despite Policy Change
Altman had previously publicly stated that OpenAI would refrain from incorporating “sex-bots” into ChatGPT, even if such features could boost user numbers. He expressed concern about preventing AI misuse that could psychologically impact individuals.
The CEO had jokingly suggested some companies might pursue Japanese anime-style sex-bots but firmly asserted OpenAI would not follow that path, prioritizing user benefit and safety instead.
New “Treat Adults Like Adults” Principle
The forthcoming update operates under the principle of “treat adult users like adults.” While ChatGPT will understand mature conversation contexts, clear rules prohibiting illegal content, abuse, or sexual violence will remain strictly enforced.
AI Personality Customization Features
Beyond content relaxation, the new ChatGPT version will introduce personality customization. Users will be able to configure responses to be warm, friendly, use emojis frequently, or converse like a friend – reflecting preferences seen in the GPT-4o model.
Altman noted that within weeks, a new version will allow users to select preferred personalities, such as friend-like personas with ample emoji use, but only when explicitly requested.
The refreshed ChatGPT will remain based on the GPT-5 model but undergo internal customization to support these new functionalities, with full public launch anticipated by December 2025.
