OpenAI Launches Automated Code Security Review Agent Codex Security

BlockBeats News, March 7th, according to official sources, OpenAI announced the launch of Codex Security, an AI-powered application security agent that can discover vulnerabilities, validate them, and propose fixes. Codex Security originated from Aardvark, a security research agent that OpenAI tested with a small number of customers last year.

The platform can analyze code repositories, perform stress testing on suspicious vulnerabilities in a sandbox environment, confirm vulnerability impact by generating Proof of Concept (PoC) attack code, and propose fixes.

OpenAI is releasing Codex Security today in a research preview version to Enterprise, Commercial, and Education customers, allowing them to use the tool for free in the first month.