GoKawiilOpenAI unveils GPT-6 Cyber, finds 81 Firefox bugs and 948 Linux exploits
OpenAI announced GPT-6 Cyber, an AI model built for vulnerability research, which uncovered 81 unknown high-severity Firefox flaws and helped patch many of them. In Linux kernel testing, the model flagged 2,308 potential issues, reproduced 1,724 of them, and built privilege-escalation exploits for 948 distinct bugs. OpenAI rated the model's cybersecurity capability as 'Critical,' up from 'High' for its predecessor, and restricted access to select customers in its Daybreak Red program.
GoKawiil's interpretation of the reporting above, not reported fact.
The jump in exploit success rate from 32.1% to 48.5% on the ExploitGym benchmark suggests AI-assisted vulnerability discovery is advancing quickly, which could reshape both offensive and defensive security work. OpenAI itself frames this as a narrowing window for defenders to patch systems before comparable AI capabilities reach attackers, implying urgency without asserting any specific attack is imminent. Restricting access to vetted customers indicates OpenAI is trying to manage dual-use risk while still commercializing the technology.
- GPT-6 Cyber found 81 new high-severity Firefox vulnerabilities, many already patched by Mozilla.
- It built working privilege-escalation exploits for 948 of 2,308 flagged Linux kernel issues.
- OpenAI classifies the model as 'Critical' risk and limits access to vetted Daybreak Red customers.
Source: bleepingcomputer.com, 2026-11-06
Published there as: “OpenAI's new GPT-6 Cyber AI finds 81 Firefox flaws, 948 Linux bugs”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.