tech
March 25, 2026
Helping developers build safer AI experiences for teens
Introducing a set of teen safety policies formatted as prompts for gpt-oss-safeguard
TL;DR
- OpenAI released prompt-based safety policies for teens to work with the gpt-oss-safeguard model.
- The policies aim to simplify the implementation of age-appropriate AI protections for developers.
- Key risk areas covered include graphic violent/sexual content, harmful body ideals, dangerous activities, roleplay, and age-restricted goods/services.
- Developed with input from Common Sense Media and everyone.ai, these policies are intended as a starting point for developers.
- The policies are released as open source to encourage collaboration and iteration within the AI safety community.
Continue reading the original article