OpenAI Introduces GPT-6 Cyber for Security Research

OpenAI has introduced GPT-6 Cyber for security research, with BleepingComputer reporting the company’s claim that it found 81 previously unknown high-severity Firefox vulnerabilities.

In separate Linux kernel research, the model flagged 2,308 potential flaws, reproduced 1,724 vulnerabilities and developed local privilege-escalation exploits for 948 distinct vulnerabilities. These are company-reported research results, not a guarantee that the model detects every software defect.

The model is not publicly available: initial access is restricted to approved Daybreak Red customers. OpenAI assigns it a Critical cybersecurity capability rating under its Preparedness Framework. Faster discovery could help defenders prioritize fixes, but human review and official security updates remain essential.

Source: BleepingComputer