OpenAI launched GPT-6 Cyber, a model for finding exploitable vulnerabilities and developing fixes, trained to stay within authorized scope.
Why it matters: A frontier model built to find and fix vulnerabilities raises dual-use stakes, making its scope controls central to its safety case.
OpenAI describes an autonomous pipeline that found, validated and generated patches for 948 Linux kernel privilege-escalation vulnerabilities.
Why it matters: Shows frontier AI automating vulnerability discovery and patching at scale, a defensive gain that also shows how cheaply exploit-finding can now be done.
Daybreak Red, which brings GPT-5.6 Cyber to approved customers, is now available in AWS GovCloud through Amazon Bedrock.
Why it matters: Puts a cyber-capable GPT model inside US government cloud for authorised users, extending controlled access to dual-use security tools.