ChatGPT stops adult mode: OpenAI postpones its premiere without a date

OpenAI has decided to indefinitely suspend the launch of the call Adult way of ChatGPT, originally planned for the first quarter of 2026. The company paused the project after intense internal discussions and red flags from employees and investors about the social, technical and commercial risks involved in a chatbot oriented to sexual content.

According to people consulted by the Financial Times, the pause is not a simple delay: the company chose to thoroughly investigate the possible effects of sexually explicit dialogue and emotional relationships between users and AI agents before moving forward. In parallel, OpenAI focuses its resources on productivity tools and integrating its services into a kind of super app.

Why does it matter now?

The decision coincides with the closure of Sora and comes at a time when regulators and societies are debating how to regulate intimate interaction with artificial intelligences. OpenAI’s action marks a precedent for corporate caution against products that can affect personal ties, security and reputation.

Internal sources report that the suspension responds to a combination of factors: ethical concerns, lack of consensus among managers, objections from some investors and technical difficulties in making the models generate that type of content without violating security regulations.

• Project status: Indefinite cancellation of launch, long-term investigation announced.
• Internal name: the function was listed in the code as Citron Mode.
• Expected requirements: age verification (over 18 years old) to activate it.
• Technical obstacles: The models were trained to block explicit sexual content, and working with sexual data requires filtering out illegal material such as incest or bestiality.
• Strategic priority: greater investment in productive tools and in the integration of ChatGPT, Codex and the browser in a single experience.

Some investors, according to sources, considered that a product with adult content offered little economic potential and increased reputational risk. There were doubts among employees about whether developing a tool designed for romantic or sentimental relationships fit with the company’s public mission: that AI benefits all of humanity.

A former high-level employee, who left the company in part over the issue, warned that AI should not replace human ties or replace real social networks. That moral concern was relevant in internal conversations.

Technical and security challenges

The teams ran into practical limitations: OpenAI models already incorporate filters and guidelines designed to reject explicit sexual content, making it difficult to reconfigure them to safely produce adult material. Furthermore, working with data sets that include sexual material requires safeguards to be established to identify and exclude illegal conduct.

OpenAI has commented that it seeks to measure and define clear boundaries, train the model to interpret contexts and ensure appropriate controls. However, the company acknowledged to its teams that there is still not enough empirical evidence on the psychological and social effects of this type of interactions.

What’s next

For now, the roadmap includes:

• A long-term study of emotional and behavioral effects.
• Internal review of policies and regulatory and business risks.
• Reinforced focus on productive products and the integration of services.

OpenAI’s pause opens relevant questions for users and regulators: how to define ethical limits in intimate experiences with AI? What controls are necessary to protect minors and prevent abuse? And to what extent should technology companies invest in products that reconcile demand and social responsibility?

At a time when public debate over emotional interaction with machines is gaining steam, OpenAI’s decision acts as a case study in how to balance innovation, security and social acceptance before launching market-sensitive features.

Similar items

Diego Soler explores technological trends and digital innovations. You will enjoy practical tutorials and direct advice to better understand artificial intelligence, cybersecurity and new technologies.