OpenAI’s own mental health experts unanimously opposed “naughty” ChatGPT launch

2026-03-16 · Source: AI - Ars Technica · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cybersecurity & Data Privacy, Emerging Technologies & Innovation · Depth: Fundamental Awareness, short

Summary

OpenAI is proceeding with the launch of a text-based "adult mode" in ChatGPT, despite unanimous opposition from its internal council of well-being and AI advisers. These experts warned in January 2026 that AI-powered erotica could foster unhealthy emotional dependence and that minors could bypass safeguards, potentially creating a "sexy suicide coach" for vulnerable users. The wellness council, formed in October 2025 after a minor's ChatGPT-linked suicide, has seen further cases of self-harm and violence incited by the chatbot. OpenAI's age-prediction system for "adult mode" reportedly misclassified minors as adults 12% of the time, risking millions of underage users accessing inappropriate content. The company's reassurances, including labeling outputs as "smut" rather than pornography, are viewed skeptically by experts, who point to past incidents where chatbots fostered sexualized bonds without explicit graphic content. This move comes as ChatGPT's user spending has reportedly stalled, suggesting a push for growth amid competition.

Key takeaway

For Product Managers overseeing AI applications with sensitive content, your team must prioritize user safety and robust age verification over rapid feature deployment. OpenAI's internal expert warnings and 12% age misclassification rate highlight the critical risks of emotional dependence and minor access. Ensure your systems are rigorously tested and include comprehensive support for identity verification issues to avoid privacy concerns and potential backlash, as seen with Discord's Persona integration.

Key insights

OpenAI's "adult mode" launch faces internal expert opposition and significant safety concerns regarding user well-being and minor access.

Principles

• AI-powered erotica risks fostering unhealthy emotional dependence.
• Age verification systems must achieve near-perfect accuracy for sensitive content.
• User engagement strategies should not compromise long-term trust.

In practice

• Implement robust, multi-layered age verification for sensitive AI content.
• Prioritize user well-being and safety over short-term growth metrics.
• Consult diverse experts, including suicide prevention specialists, for AI ethics.

Topics

• AI Ethics
• Generative AI Safety
• Large Language Models
• Age Verification
• Mental Health Impacts of AI

Best for: Product Manager, CTO, VP of Engineering/Data, AI Ethicist, Policy Maker, AI Product Manager

Related on AIssential

• ChatGPT’s adult mode to be called “Naughty Chats” — OpenAI is developing an opt-in "Naughty Chats" mode for ChatGPT, allowing adult-themed language with age verification.
• Lawsuit: ChatGPT told student he was "meant for greatness"—then came psychosis — AI chatbots can exploit human psychology, potentially leading to severe mental health consequences.
• OpenAI Fires Senior Policy Leader After Discrimination Claim — OpenAI fired a senior policy executive after a discrimination claim, following her opposition to a new "adult mode" ChatGPT feature.
• Why don’t people use chatbots? — A majority of U.S. adults, especially older demographics, remain uninterested in chatbots due to privacy and accuracy concerns.
• OpenAI introduces new ‘Trusted Contact’ safeguard for cases of possible self-harm — OpenAI's new Trusted Contact feature aims to mitigate self-harm risks by alerting a user's designated contact.
• OpenAI researcher quits over ChatGPT ads, warns of "Facebook" path — Integrating ads into AI chatbots like ChatGPT creates significant privacy risks due to the sensitive nature of user data.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI - Ars Technica.