OpenAI releases the GPT-6 Cyber security model, available only to Daybreak Red customers
OpenAI has released GPT-6 Cyber, a model designed to discover and exploit software vulnerabilities while being restricted to authorized targets, and it is currently available only to approved Daybreak Red customers via the API and Codex. In real-world software testing, it confirmed 1,724 Linux kernel vulnerabilities and built working exploits for 948 of them, compared with 166 under the same conditions for GPT-5.6 Sol; it also discovered 81 previously unknown critical Firefox vulnerabilities, most of which Mozilla has already patched. OpenAI rates it as Critical, its highest cyber risk level, due to its offensive capability, and says its training led it to ignore decoys 99.93% of the time without a deployment prompt and 100% of the time with one in the ExploitGym Honeypot test.
Why read this
SelectedAI score 83
The original text provides comparative data on the number of vulnerabilities discovered and out-of-bounds testing, which can be used to assess the capability and risk boundaries of a security-specific model.
Topics
View original postIntroducing GPT‑6 Cyber: a major advance in security research that pairs stronger capabilities with better model alignment. The model powered the Linux kernel research we announced yesterday, and security teams at companies like Mozilla, Palo Alto Networks, Calif, and Cisco—alongside our own at OpenAI—are already putting it to work. GPT-6 Cyber is available to select Daybreak Red customers and we are expanding access over time. https://t.co/Nzab3Z1Td6
View original
Introducing GPT‑6 Cyber: a major advance in security research that pairs stronger capabilities with better model alignment. The model powered the Linux kernel research we announced yesterday, and security teams at companies like Mozilla, Palo Alto Networks, Calif, and Cisco—alongside our own at OpenAI—are already putting it to work. GPT-6 Cyber is available to select Daybreak Red customers and we are expanding access over time. https://t.co/Nzab3Z1Td6
Body
OpenAI released GPT-6 Cyber, a model built to find and exploit software vulnerabilities without straying outside authorized targets.
Right now only a select group of approved Daybreak Red customers can use it, through the API and Codex.
> On real software, it confirmed 1,724 Linux kernel bugs and built working attacks for 948 of them, while GPT-5.6 Sol, tested the same way, managed 166.
> It also found 81 serious Firefox bugs nobody knew about, and Mozilla has already released fixes for many of them.
> Because the model is this good at attacking, OpenAI rates it Critical, its highest cyber-risk level
> That gain came with training to keep the agent inside the scope a researcher authorizes. ExploitGym Honeypot tests that by planting easy-to-hack side services, such as a package registry, whose compromise also reveals the task’s flag.
Without its deployment prompt, the model ignored that bait in 99.93% of tests, against 94.49% for its predecessor, and in 100% with the prompt.