OpenAI's GPT-6Cyber found 948 Linux kernel privilege-escalation bugs at $1,279 per exploit
赛博禅心 · wechat · 2026-11-06
OpenAI released GPT-6Cyber (DaybreakRed tier), rated Critical in its Preparedness cybersecurity scale. Highlights: an automated pipeline scanned the latest Linux kernel — filtering 2.8M of 40M lines reachable by unprivileged users, surfacing 2,308 suspects in 12 hours, confirming 1,724, and achieving root on default-hardened desktops (SMEP/SMAP/KASLR) for 948 unique bugs; 304 of the worst were reported to maintainers. Average API cost per exploit: $1,279. The same flow with GPT-5.6Sol found only 166 in July. The model also found 81 new high-severity Firefox bugs (41 of the 108 fixed in Firefox 159 were OpenAI's reports). In a Dragos mock power plant test, it autonomously segmented OT from IT and found two 0-days exposing breaker trip logic; new blocks now cover ICS/election-system requests. Access is split into DaybreakBlue (GPT-6Astra, lighter restrictions) and DaybreakRed (strict vetting, SSO/MFA/IP allowlists), plus a $1B fund for critical infrastructure defenders and the Codex Security scanning tool.
Related event: OpenAI Launches GPT-6 Cyber, a Security-Focused Model in Limited Release(17 posts)→
More from Models
- Researcher Jesse Yu teases "I accidentally cracked computer use" — jesselyu · 2026-11-06
- METR chart: permanent acceleration fits Anthropic's capability curve better (RMSE 1.9 vs 3.9) — scaling01 · 2026-11-06
- LLMs write evals for each other in daily "Model Fight": Gemma-4 wins, Qwen2.5-72B posts first full APEX hold — Dave_Maynor · 2026-11-06
- r-1 mini chart shows a new Pareto frontier for parsing: no more trading price for quality — JenniferHli · 2026-11-06
- METR chart: Anthropic employees self-report ~12x productivity uplift from Model 4 — scaling01 · 2026-11-06
- webAI launches open 8B model and Mac app for on-device collaborative AI agents — omarsar0 · 2026-11-06