At a Glance · AI-generated

OpenAI introduced Codex Security, a new tool designed to detect and remediate software vulnerabilities across codebases, entering research preview on March 6, 2026. The context-aware system automatically identifies security flaws and generates patches, expanding OpenAI's presence in AI-assisted code security. The launch followed OpenAI's October announcement of Aardvark, an agentic security researcher, as the company escalates its focus on developer security tools.