OpenAI Releases GPT-6 Cyber

OpenAI has released another cybersecurity-focused AI model for security research, with the release of GPT-6 Cyber.

Few industries have been so completely upended as the cybersecurity industry. Security researchers and bad actors alike have been using AI to find vulnerabilities in operating systems, platforms, and applications at an unprecedented pace. The issue has become so pronounced that the leading AI companies are restricting access to their most powerful models, only giving access to trusted companies and researchers.

OpenAI has now released another major model aimed at security research, which the company announced in a blog post:

We’re introducing GPT‑6 Cyber, a major leap in security research: more capable at finding and fixing vulnerabilities, and far more reliable staying within the boundaries researchers set. GPT‑6 Cyber delivers major gains on key cybersecurity benchmarks over GPT‑5.6 Cyber, including a 51% relative improvement on ExploitGym and a step function improvement when tested on staying within scope. It is already helping to secure real-world software, uncovering and validating thousands of exploits in security-critical codebases such as Linux and Firefox.

The new model scores significantly better than GPT-5.6 in ExploitGym tests:

On ExploitGym, which tests whether models can exploit a specified vulnerability, GPT‑6 Cyber scores 48.5%, up from 32.1% for GPT‑5.6 Cyber. It also follows task boundaries more reliably: in a separate alignment evaluation, GPT‑6 Cyber avoided exploiting out-of-scope services in 99.93% of tests without its deployment prompt, and 100% with this prompt. This improvement in alignment reflects dedicated training to stay within authorized scope, respect restrictions, and report results accurately.

The company outlined the success GPT-6 has already scored, while acknowledging the limited window available to secure software:

These capabilities are already helping secure real systems. Our researchers used GPT‑6 Cyber to develop 948 privilege escalation exploits in Linux⁠ and identify 81 previously unknown high-severity vulnerabilities in Firefox, despite extensive prior testing with frontier models. The model also helped develop candidate patches, and Mozilla has released fixes for many of the Firefox vulnerabilities⁠(opens in a new window). Security teams at Cisco, Calif, and CodeRabbit are using it for vulnerability research and code review, and we have relied on it to find and fix weaknesses in our own infrastructure⁠.

It is clear that we have a limited window to use frontier models to secure software⁠. Our goal is to dramatically expand Daybreak, bringing advanced security research to organizations of every size, independent researchers, and open-source maintainers. Mainline models such as GPT‑6 Astra already support code review, vulnerability triage, and patch development. Daybreak Blue helps verified defenders validate vulnerabilities and fixes with fewer unnecessary refusals, while Daybreak Red enables deeper exploit development and authorized security testing with GPT‑6 Cyber.

As with other similar models from Anthropic and Google, GPT-6’s access is limited to trusted partners:

GPT‑6 Cyber is initially available through the API and Codex to a select group of approved Daybreak Red customers. We will progressively expand access to all Daybreak Red customers soon, and are also making GPT‑5.6 Cyber available to all Daybreak Red customers.