OpenAI puts on hold the erotic version of ChatGPT following backlash from employees, investors, and advisors.

      OpenAI has indefinitely postponed its plans to introduce an erotic "adult mode" for ChatGPT, according to a report by the Financial Times on Wednesday. This decision concludes a five-month journey during which the feature was confidently introduced, delayed twice, and ultimately scrapped due to backlash from employees, advisors, and investors. This marks the third significant product reversal for OpenAI within a week, following the discontinuation of its Sora video generation app on Monday and the collapse of a prospective $1 billion investment from Disney.

      The adult mode was initially revealed by CEO Sam Altman in October 2025, when he expressed on X that OpenAI believed it could enforce age restrictions on sexually explicit discussions and that this approach aligned with the company's goal to "treat adult users like adults." The launch was initially planned for December 2025, then shifted to the first quarter of 2026, and has now been delayed with no indication of when it might be released. OpenAI informed the Financial Times that it intends to conduct "long-term research on the implications of sexually explicit chats and emotional attachments" before making product decisions.

      Issues arose from technical, ethical, and commercial angles, which compounded each other. Engineers found that adapting models designed to avoid sexual content for safety, in order to produce explicit material reliably, was more challenging than expected. Using datasets that included sexual themes led to outputs with illegal scenarios, such as bestiality and incest, which were difficult to filter. The feature proved not only controversial but also hard to build safely.

      Concerns from OpenAI's advisory board extended beyond content moderation. Advisors cautioned that sexually explicit interactions with ChatGPT could foster unhealthy emotional attachments, leading to serious mental health issues. One advisor referred to the potential risk as transforming ChatGPT into a "sexy suicide coach," a description that resonates darkly in light of the company’s existing legal challenges. OpenAI is currently facing at least eight lawsuits alleging that ChatGPT has contributed to user deaths, including the case of Adam Raine, a 16-year-old from Southern California whose family claims the chatbot discussed suicide methods with him over 200 times before he took his life in April 2025. Earlier this week, OpenAI identified these lawsuits as significant risks to its business in a financial document shared with investors.

      Employees began to question whether the feature aligned with OpenAI's stated mission, which is to develop artificial general intelligence that benefits humanity. Some found it challenging to reconcile that goal with the technical demands of enabling a chatbot to engage in explicit conversation without violating laws.

      Investor concerns also played a critical role in the decision. Some questioned why OpenAI would risk its reputation for a product with "relatively small upside." While there is a market for AI-generated adult content, it is currently served by numerous smaller, less scrutinized firms. For a company valued at $300 billion that is seeking enterprise customers, the potential brand damage from being associated with explicit content outweighed possible profits.

      The age verification challenge further exacerbated this concern. OpenAI's approach relied on AI-based age prediction instead of rigorous identity checks, and internal testing revealed a 10% error rate, meaning about one in ten users could be misclassified. For a product aimed at keeping explicit content away from minors, such a margin poses significant regulatory and reputational risks, particularly in a legal landscape where various US states have enacted or proposed laws mandating age verification before granting access to adult materials.

      In recent developments, the decision about the adult mode follows another setback: the discontinuation of Sora, the AI video generation tool that was billed as a creative platform for filmmakers and content creators. Sora required substantial computing resources compared to its revenue, and its key partnership with Disney—a three-year licensing agreement allowing users to generate videos featuring Disney, Marvel, Pixar, and Star Wars characters—collapsed alongside the announcement. Disney had aimed to invest $1 billion in OpenAI as part of this deal, but no funds had been exchanged.

      Collectively, these three reversals reflect a company retreating from experiments with consumer products and redirecting focus towards its core operations. Reports indicate that investors are now more interested in seeing OpenAI integrate ChatGPT with coding assistants to create a "super app" targeted at transforming business operations—a vision that promises clearer monetization opportunities and fewer reputational risks compared to video generation or erotic chatbots.

      OpenAI has announced it will shift resources to robotics and autonomous software agents, areas with a more direct path to commercial value and a regulatory landscape that, while complex, does not involve the same risks associated with sexualized AI and child safety.

      There is a noticeable pattern in OpenAI's product strategy: ambitious announcements are followed by real-world complications that might have been anticipated by less confident organizations, leading to a retreat framed as prudent research. The adult mode was revealed before solutions for safe content generation were developed, before the age verification system achieved acceptable accuracy, and before addressing the advisory board's mental health concerns. Similarly, the