Skip to main content
AI RADARModels
1•11/05/2026, 22:33•2 min read

OpenAI introduces GPT-6 Cyber, an AI model that found 948 vulnerabilities in the Linux kernel

OpenAI introduces GPT-6 Cyber, an AI model that found 948 vulnerabilities in the Linux kernel - 1
OpenAI introduces GPT-6 Cyber, an AI model that found 948 vulnerabilities in the Linux kernel - 2
OpenAI introduces GPT-6 Cyber, an AI model that found 948 vulnerabilities in the Linux kernel - 3
Executive TL;DR30s read

OpenAI has introduced GPT-6 Cyber, a specialized AI model that discovered 948 vulnerabilities in the Linux kernel and more than 80 critical flaws in Firefox.

OpenAI has introduced GPT-6 Cyber, a specialized AI model designed to find software vulnerabilities. According to the company, the model discovered 948 vulnerabilities in the Linux kernel.

Researchers built an autonomous pipeline around the model to search for vulnerabilities.

On the ExploitGym benchmark, GPT-6 Cyber scored 48.5%. By comparison, GPT-5.6 Cyber scored 32.1%, representing a relative improvement of about 51%.

Beyond the Linux kernel, the model also helped uncover more than 80 new critical vulnerabilities in Firefox. It is being used by Mozilla, Cisco, Palo Alto and other organizations.

OpenAI has also started measuring the model's tendency to go beyond the boundaries of a task. The company says that in the new version this behavior has significantly decreased compared with GPT-5.6.

The model is currently available only in Daybreak Red.

Why it matters

Analysis

This shows how AI can accelerate security audits of large open-source projects, while also increasing the need for oversight of autonomous tools.

Discuss in community

Share your questions and insights with developers

Join Telegram
+5 Points
← Back to news feed

Related in Ecosystem

Practical guides, core terminology, and agent skills