OpenAI has launched Codex Security, an AI-powered security agent that can identify, validate, and propose fixes for vulnerabilities.
This feature is currently available in a research preview to select customers, offering free usage for the next month.
According to OpenAI, Codex Security enhances security measures by detecting and fixing vulnerabilities on a large scale.
The AI company reports that Codex Security has already identified critical and high-severity findings in various open-source projects.
Furthermore, OpenAI states that Codex Security has improved its precision and reduced false positives over time.
The application security agent works by analyzing the project’s security structure, identifying vulnerabilities, and proposing fixes that align with the system behavior.
Recently, Anthropic also introduced Claude Code Security, another tool aimed at scanning codebases for vulnerabilities and suggesting patches.
