OpenAI, which has been making rapid strides in the field of generative AI, seems to be re-evaluating the boundaries of its product line recently. Following its recent announcement...Close Sora video generation toolLater, OpenAI confirmed again that it would "indefinitely" abandon its planned launch.Adult Porn Chatbot Project (Internal code name: "Citron mode").
This decision was driven by investor concerns, employee anxieties, and the technical bottleneck of completely eliminating illegal content during model training. OpenAI stated that it will focus its resources on core productivity tools such as coding assistants, officially marking the end of this "adult turnaround" experiment.
The tug-of-war between technology and ethics: The truth behind the difficult birth of "Citron mode"
OpenAI had previously revealed in October 2025 that it was considering relaxing restrictions on adult content and planned to launch related features by the end of that year. However, the delay and eventual shelving of the plan highlights two major challenges in the development process:
• Filtering mechanism malfunction:Sources revealed that when the development team retrained the model that originally avoided pornographic content, they found it difficult to accurately filter out illegal behaviors such as bestiality or incest, which poses a huge brand risk to OpenAI, which prides itself on safety and compliance.
• Age identification error:Although OpenAI has implemented automatic age detection technology, its error rate is still over 10%. This means that with hundreds of millions of users worldwide, a large number of minors may be exposed to restricted content, potentially leading to a series of legal lawsuits.
Internal opposition: AI should not replace human emotional connection
This plan has caused a major upheaval within OpenAI. The report indicates that a senior employee even resigned as a result, stating bluntly, "AI shouldn't replace your friends or family; what humans need is genuine connection."
Furthermore, rival xAI's Grok model has been embroiled in controversy for generating deepfake nude photos of real people and children, which has also unsettled OpenAI's investors. Although CEO Sam Altman initially believed it was "safe to relax restrictions," the company ultimately decided to put on the brakes due to a lack of sufficient empirical research on the impact of AI attachment on mental health.
Strategic Transformation: Abandoning "Side Quests" and Focusing on Productivity
OpenAI has stated that it will currently focus on tools that can improve human efficiency, and has defined Sora video generation and adult chatbots as "side quests." This indicates that in 2026, with expensive computing resources and increasingly stringent regulations, OpenAI is attempting to scale back its efforts and refocus on improving the fundamental capabilities and commercial applications of the GPT series.
