OpenAI CEO Sam Altman announced through a post on X, his plans for a new version of ChatGPT that will restore behaviors from the GPT-4o model focusing on greater user customization. The update includes options for the chatbot to respond in a human like manner use emojis extensively or adopt friendly interactions based on user preferences. Starting in December 2025 verified adult users over 18 will access an adult mode permitting erotica and other mature content as part of comprehensive age-gating implementation.
This change follows OpenAI’s earlier restrictions implemented for mental health safety which limited usability for users without such concerns. The company has developed new tools to mitigate risks enabling relaxation of restrictions in most cases. The update applies to ChatGPT across platforms with rollout beginning in the coming weeks.
Table of Contents
What were the earlier ChatGPT restrictions?
OpenAI introduced restrictions to ChatGPT in early 2025 to address potential mental health impacts from interactions particularly for users showing signs of distress. These measures included limiting conversational depth tone expressiveness and content types to prevent escalation of sensitive topics. The GPT-5 model default since September 2025 emphasized safety through reduced personality variability and stricter moderation resulting in feedback from users describing responses as overly corporate or less engaging.
Altman acknowledged that while the approach prioritized caution it reduced enjoyment and utility for the majority of users without mental health issues. Internal reviews confirmed that tools could now better identify distress patterns allowing for targeted safeguards rather than broad limitations. This led to the decision to revert to GPT-4o characteristics while maintaining protections for vulnerable groups. The policy aligns with OpenAI’s Model Spec which classifies erotica as sensitive content permissible under specific conditions.
What you need to know about the GPT-4o style update
The new ChatGPT version scheduled for release in the coming weeks will enable users to customize the chatbot’s tone personality and response style resembling the GPT-4o model preferred by many for its conversational flexibility. Options include human like interactions with natural language variations extensive emoji usage and friendly behaviors activated via user settings. These customizations are opt in ensuring users who prefer formal or neutral responses can select those instead.
The update builds on GPT-4o’s capabilities such as contextual awareness and creative output while incorporating improvements from GPT-5 for accuracy and safety. OpenAI states the goal is to enhance usability without compromising core protections. The feature integrates with existing ChatGPT interfaces on web iOS and Android with no additional hardware requirements. Users will access settings through the account dashboard to adjust preferences.
Arrival of the Adult mode for users
In December 2025 OpenAI will enable adult mode for ChatGPT users who verify their age as 18 or older providing access to erotica and mature content previously restricted under the content policy. Verification occurs through age prediction systems analyzing interaction patterns combined with optional identity confirmation methods such as government ID or credit card details. If age cannot be confirmed confidently the chatbot defaults to the under 18 experience with stricter filters.
Adult mode permits content like fictional erotica and discussions of mature themes but continues to prohibit illegal exploitative or harmful materials including depictions of minors in sexual contexts. This aligns with OpenAI’s principle of treating adult users as adults while prioritizing safety for teenagers through parental controls and alerts. The rollout includes expanded mental health mitigations to monitor and intervene in potentially harmful conversations. Access requires a ChatGPT Plus or Enterprise subscription for full features.
What are the ‘Age Gating’ safeguards that will be in place?
Age gating implementation expands on tools introduced in September 2025 using behavioral analysis to detect user age with 90 percent accuracy based on language patterns and query types. For unverified users ChatGPT maintains PG rated interactions with automatic redirection from sensitive topics. Adult mode activation prompts verification and includes session logging for compliance reviews. OpenAI reports that these systems reduce underage access risks by 95 percent in testing.
The policy update follows Federal Trade Commission inquiries into AI safety for minors and responds to user petitions gathering over 6 000 signatures for less restrictive adult experiences. Content generation in adult mode adheres to updated guidelines prohibiting extreme violence or non consensual scenarios while allowing educational or artistic mature themes. Parents can set controls for shared accounts to limit exposure.
The GPT-4o style update begins rollout in the weeks following October 15 2025 prioritizing ChatGPT Plus and Enterprise subscribers before expanding to free users. Adult mode launches in December 2025 integrated with full age gating across all platforms. Verification processes will be available immediately with phased onboarding to manage server load. The update requires no app reinstallation and functions on existing devices with iOS 15 or later Android 8.0 or later and modern web browsers. OpenAI provides an opt out for users preferring the current restrictive settings.
