# OpenAI Releases Safety Report for New o3-mini Model

OpenAI has published a system card detailing the safety measures implemented for its new o3-mini model, marking another step in the company's transparency efforts around AI safety.

The report, announced via the company's official Twitter account, covers three key areas: comprehensive safety evaluations, external red teaming exercises, and assessments under OpenAI's Preparedness Framework. System cards are technical documents that explain how AI models are tested for potential risks before public release.

The o3-mini model represents OpenAI's latest addition to its reasoning model lineup, designed to be a more compact and efficient version compared to larger models. By releasing this safety documentation, OpenAI is providing researchers, policymakers, and the public with insight into how the company identifies and mitigates potential risks such as harmful content generation, bias, and security vulnerabilities.

External red teaming—where