OpenAI is launching "Age Predictions" for ChatGPT users across the European Union, enhancing safety by estimating if users are under 18. This new feature assesses user-provided age and account activity patterns to ensure content moderation.
Enhanced Content Control
The Age Predictions system will limit exposure to sensitive materials for users identified as possibly under 18. This includes content such as graphic violence, risky challenges, and unhealthy beauty trends. The tool aims to protect younger users by tailoring content accessibility based on inferred age indicators.
Identity Verification and Parental Options
If users believe they've been incorrectly classified, they can verify their age using Persona, a secure identity service. Parents gain access to control settings like establishing "quiet hours," limiting what the AI remembers about their child, adjusting content for model training, and receiving alerts if potential distress signals are noticed.
Rollout Schedule and Additional Features
OpenAI intends to deploy Age Predictions gradually in the European Union over the coming weeks, refining the system for better accuracy. This development follows the introduction of ChatGPT Go, a premium subscription model with higher usage limits, and a pilot program for chat advertising that aims to deliver relevant, non-intrusive ads.
