OpenAI's adult mode will allow erotic text but blocks explicit image, audio, and video generation
OpenAI confirmed its forthcoming "adult mode" will permit text-based erotic conversations in ChatGPT but explicitly block generation of pornographic images, audio, and video. The feature, first announced by CEO Sam Altman in October 2024, has been delayed multiple times—most recently in March 2025—as the company grapples with safety concerns including a 12% error rate in age verification systems.
OpenAI's Adult Mode Will Allow Erotic Text but Blocks Explicit Image, Audio, and Video Generation
OpenAI has clarified that its forthcoming "adult mode" will permit users to engage in erotic text conversations with ChatGPT, but will strictly prohibit generation of pornographic images, audio, or video. According to reporting from The Wall Street Journal, an OpenAI spokesperson described the feature as producing "smut rather than pornography."
CEO Sam Altman first proposed the adult mode concept in October 2024, stating OpenAI wanted to "treat adult users like adults." The company originally targeted a January 2026 launch but has delayed the rollout multiple times, with the most recent postponement occurring in March 2025 as OpenAI cited "focus on work that is a higher priority for more users."
Safety Concerns Outweighed
OpenAI proceeded with adult mode development despite unanimous warnings from its eight-member Council on Wellbeing and AI. Council researchers expressed concerns that AI-generated erotica could foster unhealthy emotional dependence on ChatGPT, with one member specifically warning of the risk of creating a "sexy suicide coach" in reference to documented cases where users have taken their own lives after becoming emotionally attached to the chatbot.
The age verification challenge proved substantial. OpenAI's age prediction technology misidentified underage users as adults approximately 12% of the time during testing. With roughly 100 million teens using ChatGPT weekly, this error rate would potentially grant millions of minors access to erotic content. OpenAI told The Journal its algorithm performs to industry standards and acknowledged no system will be completely foolproof.
Implementation Constraints
The explicit restriction on image, audio, and video generation represents a significant limitation compared to ChatGPT's broader multimodal capabilities. The text-only constraint appears designed to reduce the likelihood of generating photorealistic or deepfake-style explicit material, though OpenAI did not publicly elaborate on this rationale.
The company has not disclosed a confirmed launch date for adult mode, and no pricing or access restrictions for the feature have been announced.
What This Means
OpenAI is balancing commercial expansion into adult content against documented safety risks and regulatory pressure. The text-only limitation and repeated delays suggest internal conflict between product teams and safety advisors. The 12% age verification error rate—if accurate—highlights fundamental technical challenges in content access control at scale. For competitors and regulators, this case demonstrates that even well-resourced companies struggle to implement meaningful safeguards for sensitive user segments.
Related Articles
Google expands Gemini Android overlay menu with six new tools accessible without opening app
Google has expanded the Gemini overlay plus menu on Android to include six tools: Videos, Music, Canvas, and Guided Learning join the existing Images and Personal Intelligence options. The update, rolling out in Google app version 17.32, allows users to access most Gemini features from anywhere on Android without opening the full app.
Trail of Bits and OpenAI's Daybreak initiative produce 64 pull requests across 19 open-source projects in one week using
Trail of Bits launched Patch the Planet, a security initiative using OpenAI's GPT-5.5-Cyber model to find and fix bugs in critical open-source projects. The first week produced 64 pull requests and 51 issues across 19 projects including cURL, Python, PyPI, and Sigstore, with 37 patches already merged.
Mistral AI adds Deep Research agent, voice mode with Voxtral model to Le Chat
Mistral AI has released a major update to Le Chat, adding a Deep Research agent that generates structured research reports, a new voice input model called Voxtral, and Projects for organizing conversations. The update also includes multilingual reasoning powered by Mistral's Magistral model.
Tencent tests AI assistant Xiaowei in WeChat's 1.4 billion user base
Tencent is testing an AI assistant called Xiaowei in Weixin, the Chinese version of WeChat, which has over 1.4 billion monthly active users combined with WeChat. Users can interact with Xiaowei through text or voice, communicate with friends, and launch mini-programs within the app.
Comments
Loading...