**[OpenAI Launches AI Agent Security Tool Codex Security]** Foresight News reports that OpenAI has announced the launch of Codex Security, an application security agent designed to help users protect codebases by identifying vulnerabilities, verifying their validity, and providing reviewable and actionable remediation suggestions. This feature enables teams to focus on critical vulnerabilities and accelerate code delivery. Codex Security is the successor to last year’s Aardvark, with significant improvements in signal quality, severity accuracy, and reduced false positive rates, ensuring that identified vulnerabilities better align with real-world risks. Additionally, Codex Security will be offered free of charge to ChatGPT Enterprise, Business, and Edu customers as a research preview version over the next month.