The End of Lying About Your Age: ChatGPT Now “Guesses” Your Age to Block NSFW Content

OpenAI has unveiled a sophisticated AI-driven age prediction mechanism within ChatGPT that moves beyond user-provided birth dates by analysing “behavioral patterns” to determine a user’s age, with safeguards that trigger additional content filters if a user is deemed under 18. The feature operates globally, and a prior “Teen Edition” of ChatGPT allowed guardians to manage memory, chat histories, and set “prohibited intervals” to curb late-night use.

If the system flags a user as under 18, it enforces strict safety protocols that can limit exposure to erotica, violence, or self-harm, while the age-determination logic considers general themes discussed and temporal usage patterns; even falsified ages may be overridden if dialogue aligns with adolescent characteristics.

If misclassification occurs for an adult, an appeals process is available via a “Verification” option that routes users to a third‑party partner, Persona, for identity validation, requiring a live selfie and government‑issued ID. According to OpenAI, Persona handles the documents and photos, with OpenAI only receiving age data and verification status, and Persona deleting image data within seven days. 22 January 2026