OpenAI has indefinitely shelved plans to introduce an erotic “adult mode” to ChatGPT, as reported by the Financial Times. This decision follows a five-month period where the feature faced delays and eventual cancellation due to dissent from staff, advisors, and investors. This marks OpenAI’s third major product reversal within a week, following the closure of its Sora video generation app and the collapse of a $1 billion investment from Disney.
The adult mode was initially announced by CEO Sam Altman in October 2025, anticipating that sexually explicit conversations could be age-restricted and align with the principle to treat adult users like adults. It was first set for a December 2025 release, then postponed to early 2026, and now lacks a release timeline. OpenAI intends to conduct extensive research on the impacts of sexually explicit chats before making any product decisions.
The problems faced were technical, ethical, and commercial. Technically, engineers struggled to train models to produce explicit content while avoiding illegal materials like bestiality and incest. Advisors raised concerns about unhealthy emotional attachments and mental health risks, with existing lawsuits adding pressure. One case involves a 16-year-old’s suicide allegedly linked to chat interactions with ChatGPT. Staff questioned whether the feature aligned with OpenAI’s mission of building AI for the benefit of humanity.
Investors questioned the economic viability versus the reputational risks involved. OpenAI’s reliance on AI-based age prediction, with a 10% error rate, posed significant concerns in a legal landscape requiring age verification for adult content.
OpenAI’s decision follows the discontinuation of Sora, an AI video tool, which saw a collapsed partnership with Disney when the latter withdrew a planned $1 billion investment. These shifts indicate a refocus on the core business. Investors now prefer combining ChatGPT with coding assistants for a “super app” aimed at business transformation, offering clearer monetization pathways without the reputational risks of video generation or adult chatbots.
OpenAI will pivot toward robotics and autonomous software agents, which present more direct commercial value and regulatory clarity. The company’s pattern appears to involve ambitious announcements, encountering real-world challenges, and strategically retreating to frame reversals as prudent research steps. Recent reversals suggest internal and external pressures are aligning OpenAI’s operations towards commercially and ethically viable paths, although the consistency of this mechanism remains to be seen with future announcements.