OpenAI launches GPT-6 Cyber to find and fix security flaws

OpenAI announced GPT-6 Cyber, a model built to find software vulnerabilities and help write patches.

November 5, 2026 · Launch

OpenAI announced GPT-6 Cyber, a model built to find software vulnerabilities and help write patches. It is going first to a limited group of customers in OpenAI’s Daybreak Red program for security testing, with extra identity checks, closer monitoring and more frequent check-ins on high-risk requests. OpenAI says the model found at least 81 new vulnerabilities in Firefox, built exploits for 948 Linux kernel vulnerabilities and found a critical flaw in the Brotli compression library, all reported and patched. These are OpenAI’s own figures. No public pricing was given.

Why it matters to you: AI is now finding security holes faster than people can. That should mean quicker fixes for the apps you use, and it is another reminder to install updates promptly at home and at work.

Try this: Do a quick security check-up for your home or small business.

I run [a small business with X staff / a household] and we use [list devices and apps]. Give me a 10-step security checklist, in plain language and in priority order, covering updates, passwords, two-factor sign-in, backups and phishing.

Source: CyberScoop

More from Today in AI

All of Today in AI. Where AI is heading.

Related Power of AI pages

Keep reading with Start here, everyday uses, the tool guide, the writing workshop, and sources and standards.