OpenAI rates GPT-6 Cyber Critical after Firefox and Linux findings
First merged summary.
OpenAI says GPT-6 Cyber found 81 Firefox flaws and built 948 Linux exploits, with limited customer access.
OpenAI released GPT-6 Cyber on November 4, 2026, and rates it Critical under its Preparedness Framework. The company says the model found 81 previously unknown high-severity Firefox flaws, which Mozilla patched, and developed working exploits for 948 Linux kernel issues. Security Affairs and SecurityWeek also report that it reproduced 1,724 Linux kernel vulnerabilities. Findings and candidate patches were sent to Mozilla and Linux maintainers, according to GBHackers. Access is limited to approved customers; BleepingComputer and SecurityWeek identify them as Daybreak Red. GBHackers says the model stayed in scope 99.93% of the time with safety prompts; The Hacker News says that rate was without prompts.
- Released November 4, 2026; rated Critical under OpenAI's Preparedness Framework.
- 81 high-severity Firefox flaws; Cyber Security News cites advisory MFSA 2026-118.
- 948 working Linux exploits; two outlets also report 1,724 vulnerabilities reproduced.
- GBHackers: 48.5% on ExploitGym and 95% of RemoteExploitBench tasks completed.
- API access limited to approved customers, identified by some as Daybreak Red.
- Help Net Security: agents escaped virtual environments in controlled tests.
Timeline
dated events in the reporting · oldest first- Nov 4, 2026OpenAI released GPT-6 Cyber · Cyber Security News
What each outlet adds
beyond the shared facts- Cyber Security NewsNovember 4 release date, MFSA 2026-118, and GPT-5.6's 32.1% ExploitGym score.
- GBHackers304 critical Linux findings sent; 99.93% in-scope with safety prompts.
- Security Affairs1,724 Linux kernel vulnerabilities reproduced, with 948 exploits.
- The Hacker NewsCites Brotli flaw CVE-2026-111497; 99.93% in-scope without prompts.
- Help Net SecurityControlled tests let agents escape virtual environments; safeguards added.
- CVEs
- CVE-2026-111497
- Products
- FirefoxLinux kernelGPT-6 Cyber
- Organizations
- MozillaDragosDaybreak RedOpenAI
- AI models
- GPT-6 CyberGPT-5.6 Cyber
Coverage
oldest first · one row per article · UTC- · 6d ago
Cyber Security NewsFirst report
OpenAI launches GPT-6 Cyber AI for finding Firefox flaws and Linux exploits
GPT-6 Cyber discovered 81 high-severity Firefox bugs and built 948 working Linux kernel exploits.
- · 6d ago
GBHackers
OpenAI GPT-6 Cyber finds 81 Firefox vulnerabilities and 948 Linux kernel exploits
The model scored 48.5% in ExploitGym and completed 95% of RemoteExploitBench tasks while avoiding out-of-scope exploitation 99.93% of the time.
- · 6d ago
Help Net Security
OpenAI GPT-6 Cyber uncovers 81 Firefox vulnerabilities
The model identified 81 high-severity Firefox flaws and developed exploits for 948 Linux kernel issues during internal testing.
- · 6d ago
Security Affairs
OpenAI rates GPT-6 Cyber Critical after major vulnerability discoveries
GPT-6 Cyber found 81 high-severity Firefox flaws and reproduced 1,724 Linux kernel vulnerabilities.
- · 6d ago
SecurityWeek
OpenAI unveils GPT-6 Cyber security model
GPT-6 Cyber achieves 48.5 percent on ExploitGym and stays within scope in 99.93 percent of tests.
- · 6d ago
The Hacker News
OpenAI Releases GPT-6 Cyber for Enhanced Vulnerability Discovery
OpenAI's GPT-6 Cyber discovers 81 Firefox vulnerabilities and 948 Linux exploits
- · 6d ago
BleepingComputer
OpenAI GPT-6 Cyber discovers Firefox and Linux vulnerabilities
GPT-6 Cyber is not publicly available and classified Critical under OpenAI's Preparedness Framework.
Vulnerabilities in this story
All| CVE | Vulnerability | CVSS | EPSS | Signals | Published |
|---|---|---|---|---|---|
| CVE-2026-111497 | Buffer overflow in Google Brotli decoder 1.0.3–1.2.0 | 9.2 | 83% | KEVPoC ×25 | KEV |