GPT OSS Safeguard 120B is OpenAI’s open-weight safety model for content moderation and guardrail enforcement. It helps teams evaluate content against custom safety policies and build scalable, adaptable protections for production AI applications.