AI security infographic: Aardvark icon scanning code, arrows to detected vulnerabilities and patches; protective shields for automation, with "Secure Your Code" banner over a software repository.

OpenAI has launched Aardvark, an AI agent powered by GPT-5, designed to autonomously detect software vulnerabilities and generate fixes. Announced on October 29, 2025, it addresses the surge in CVEs by scaling human-like analysis for developers and security teams.

AI security infographic: Aardvark icon scanning code, arrows to detected vulnerabilities and patches; protective shields for automation, with "Secure Your Code" banner over a software repository.AI security infographic: Aardvark icon scanning code, arrows to detected vulnerabilities and patches; protective shields for automation, with "Secure Your Code" banner over a software repository.

Aardvark analyzes codebases, scans commits, validates flaws in a sandbox, and proposes patches using Codex. It integrates with GitHub, detected 92% of flaws in tests, and has led to 10 CVE disclosures. Unlike traditional tools, it uses LLM reasoning for deeper insights.

AI security infographic: Aardvark icon scanning code, arrows to detected vulnerabilities and patches; protective shields for automation, with "Secure Your Code" banner over a software repository.AI security infographic: Aardvark icon scanning code, arrows to detected vulnerabilities and patches; protective shields for automation, with "Secure Your Code" banner over a software repository.

Available in private beta, OpenAI offers pro-bono scanning for non-commercial projects. This tool democratizes security, reducing exploitation risks and fostering safer software development.


NPAV offers a robust solution to combat cyber fraud. Protect yourself with our top-tier security product, FraudProtector.net