OpenAI Tries to Balance Between Privacy, Freedom & Teen Safety

Sam Altman outlined OpenAI’s principles, striking a balance between privacy, freedom, and teen safety in the use of AI. He emphasized that AI conversations, often profoundly personal, warrant protections comparable to those of doctor-patient or lawyer-client privilege, with robust security measures ensuring privacy even from employees. Exceptions include automated monitoring for major risks, such as threats to life, harm, or large-scale cyber incidents. For adults, OpenAI seeks to maximize freedom with models that won’t default to flirtation or harmful instructions, but will support such content in safe, consensual, or creative contexts. Teens, however, receive stricter safeguards: age prediction systems, ID checks when needed, bans on flirtatious or self-harm discussions, and parental or authority alerts in cases of suicidal ideation. Altman acknowledged trade-offs, aiming for transparency in the face of challenging choices.

OpenAI, Sam Altman – Teen safety, freedom, and privacy – 16th September 2025