Skip to content
NewA calmer, faster ZeroHour
StoryModel releasehigh

OpenAI Releases GPT-6 Cyber AI Model

14 sources · 14 articlesFirst reported Updated ()1
What's new

November 6: BleepingComputer reports GPT-6 Cyber not publicly available and expanding access to approved organizations.

Merged summary · x-ai/grok-4.7 · rewritten as coverage arrivesSummaries by Vani AI

OpenAI launches restricted GPT-6 Cyber AI that scored 48.5% on ExploitGym while staying in scope 99.93% of tests.

OpenAI released GPT-6 Cyber on November 4, 2026, a restricted-access AI model rated Critical under its Preparedness Framework. The model outperformed GPT-5.6 Cyber and GPT-6 Astra on ExploitGym with 48.5% while maintaining 99.93% in-scope behavior. It identified 81 previously unknown high-severity Firefox flaws that Mozilla began patching and developed 948 working Linux kernel privilege-escalation exploits, including 588 novel. Access is initially limited to approved Daybreak Red customers with plans to expand.

Key facts
  • 48.5% on ExploitGym versus 32.1% for GPT-5.6 Cyber
  • 99.93% in-scope behavior on ExploitGym tests
  • 81 new Firefox high-severity flaws reported to Mozilla
  • 948 Linux kernel exploits developed, 588 previously unknown
  • Rated Critical in OpenAI Preparedness Framework
  • Initial access to approved Daybreak Red customers
Explain this AI news with Vani AIWhat was announced, what is new and why it matters, in a new chat

Timeline

dated events in the reporting · oldest first
  1. Nov 4, 2026OpenAI releases GPT-6 Cyber · Cyber Security News
  2. Nov 5, 2026Model scores 48.5% on ExploitGym and stays in scope 99.93% · The Register
  3. Nov 5, 2026Mozilla begins patching 81 new Firefox flaws · The Register
  4. Nov 5, 2026OpenAI rates GPT-6 Cyber Critical · Security Affairs
  5. Nov 6, 2026Access expands to approved organizations · BleepingComputer

What each outlet adds

beyond the shared facts
  • The Register0.07% boundary tests exceeded scope; promises stays within rules of engagement
  • The VergeOut-of-bounds fell below 1% from 5%; better at not hacking other companies
  • CyberScoopValidated critical RCE flaw in Brotli; partnership with Dragos
  • CSO Online10 novel exploitation techniques; remediation rate under 25% on Patch the Planet dashboard
  • The DecoderReproduced 1,724 Linux kernel vulnerabilities; campaign cost $1,213,430 or $1,279 per exploit
CVEs
CVE-2026-111497
Vendors
OpenAIMozilla
Products
DaybreakDaybreak RedDaybreak BlueFirefoxLinux kernelCodexChatGPTAPI PlatformGPT-5.6 CyberBrotliGPT-6 Cyber
Organizations
MozillaCiscoCodeRabbitCalifDragosDaybreak RedLinux maintainersCloudflareNoma SecurityOpenAI
AI models
GPT-6 CyberGPT-5.6 CyberGPT-6 AstraGPT-6.1 AstraGPT-6

Coverage

oldest first · one row per article · UTC
  1. · 4d ago OpenAI NewsFirst report

    OpenAI introduces GPT-6 Cyber for scoped vulnerability research

    OpenAI reports 48.5% on ExploitGym, up from 32.1%, and 99.93% in-scope behavior without the deployment prompt.

  2. · 3d ago The Register · Security

    OpenAI releases GPT-6 Cyber, rated Critical for cyber capability

    Access starts with approved Daybreak Red customers after OpenAI said the model stayed in scope over 99 percent of the time.

  3. · 3d ago The Verge · AI

    OpenAI launches GPT-6 Cyber, rated Critical on its framework

    OpenAI says out-of-bounds behavior fell below 1 percent from about 5 percent, and the ExploitGym score rose to 48.5 percent.

  4. · 3d ago CyberScoop

    OpenAI debuts GPT-6 Cyber for limited Daybreak Red customers

    OpenAI says the model found 81 novel Firefox bugs and built exploits for 948 Linux kernel flaws, all since patched.

  5. · 3d ago CSO Online

    OpenAI's GPT-6 Cyber finds 81 unknown high-severity Firefox bugs and 588 kernel flaws

    OpenAI claims 10 of the model's exploitation techniques were novel; its dashboard shows under 25% of found flaws remediated.

  6. · 3d ago TechCrunch · AI

    OpenAI releases cybersecurity model for Daybreak

    OpenAI launched ChatGPT Cyber for its Daybreak program, enabling frontier models for vulnerability hunting while researchers found issues in Linux and Firefox.

  7. · 3d ago The Decoder

    OpenAI releases GPT-6 Cyber model for vulnerability hunting

    GPT-6 Cyber finds vulnerabilities in Firefox and Linux kernel with improved safety; initial access limited to select customers.

  8. · 3d ago Cyber Security News

    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.

  9. · 3d 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.

  10. · 3d 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.

  11. · 3d 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.

  12. · 3d 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.

  13. · 3d 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

  14. · 2d 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
CVEVulnerabilityCVSSEPSSSignalsPublished
CVE-2026-111497
Buffer overflow in Google Brotli decoder 1.0.3–1.2.0
9.283%KEVPoC ×25
KEV