# OpenAI Releases Open-Weight Safety Models for Developers

OpenAI announced gpt-oss-safeguard, a new set of open-weight reasoning models designed specifically for safety classification. The company shared the news via its official Twitter account.

**What's New**

Unlike OpenAI's typical closed models, these safety classification tools are released with open weights, meaning developers can download and run them independently. The models help identify potentially harmful content and policy violations in AI applications.

**Key Feature**

The standout capability is customization. Developers can now apply and iterate on their own safety policies rather than relying solely on OpenAI's predefined rules. This gives teams flexibility to tailor content moderation to their specific use cases and community standards.

**Why It Matters**

This release addresses a growing need in AI development: scalable, customizable safety tools. As more companies build AI-powered applications, they need