Live Feeds
● LIVE Updated 1h ago Β· 14 sources tracked

OpenAI Launches GPT-6 Cyber After AI Helps Find 81 Firefox Flaws, 948 Linux Exploits

OpenAI released GPT-6 Cyber on Thursday to a limited group of vetted customers. The model is rated "Critical" under the company's Preparedness Framework for cybersecurity capabilities. During testing, it discovered 81 previously unknown high-severity Firefox vulnerabilities and developed working exploits for 948 distinct Linux kernel flaws. OpenAI says the model includes safeguards to stay within authorized rules of engagement and is designed to help researchers find, validate, and fix software flaws before attackers can exploit them.

πŸŽ™οΈ

Listen to Live Briefing

Real-time synthesized voice briefing Β· Live Feeds Desk

⏱ ~3 min
Speed:
RSS Source map (14)
⚑ Key Developments & Real-Time Context
Text size:
  • βœ“ OpenAI launched GPT-6 Cyber on Thursday to a limited group of vetted customers.
  • βœ“ The model is rated "Critical" under OpenAI's Preparedness Framework for cybersecurity capabilities.
  • βœ“ GPT-6 Cyber discovered 81 previously unknown high-severity Firefox vulnerabilities.
  • βœ“ The model developed working exploits for 948 distinct Linux kernel flaws.
πŸ›‘οΈ Source Corroboration: 14 independent reporting domains (88% confidence) ⏱ Read time: ~2 min

What changed

OpenAI debuted GPT-6 Cyber, a specialized cybersecurity model that uncovered 81 Firefox bugs and 948 Linux exploits during pre-release testing, and made it available to vetted customers under a Critical risk rating.

Live updates

  1. OpenAI launches GPT-6 Cyber after model finds 81 Firefox flaws, 948 Linux exploits

    OpenAI released GPT-6 Cyber on Thursday to a limited group of vetted customers. The model is rated "Critical" under the company's Preparedness Framework for cybersecurity capabilities. During testing, it discovered 81 previously unknown high-severity Firefox vulnerabilities and developed working exploits for 948 distinct Linux kernel flaws. OpenAI says the model includes safeguards to stay within authorized rules of engagement and is designed to help researchers find, validate, and fix software flaws before attackers can exploit them.

    Why it matters

    The launch signals a shift toward AI agents that can autonomously discover and weaponize vulnerabilities at scale. A "Critical" rating under OpenAI's own framework indicates the model poses significant risk if misused, prompting restricted access. The Firefox and Linux findings demonstrate practical offensive capability across widely deployed software, raising questions about disclosure timelines and defensive patching capacity.

    What is confirmed

    • OpenAI launched GPT-6 Cyber on Thursday to a limited group of vetted customers.
    • The model is rated "Critical" under OpenAI's Preparedness Framework for cybersecurity capabilities.
    • GPT-6 Cyber discovered 81 previously unknown high-severity Firefox vulnerabilities.
    • The model developed working exploits for 948 distinct Linux kernel flaws.
    • OpenAI says the model includes safeguards to stay within authorized rules of engagement.

    Still unconfirmed

    • The model uncovered 588 unknown Linux vulnerabilities in addition to the 948 exploits.
    • GPT-6 Cyber can validate and fix software flaws autonomously before attackers exploit them.

    What to watch next

    • Disclosure and patching timelines for the 81 Firefox and 948 Linux vulnerabilities
    • Expansion of GPT-6 Cyber access beyond the initial vetted customer group
    • Independent verification of the model's "rules of engagement" safeguards in real-world use
    Sources used for this update (16)
    1. www.youthassembly.kr β€” [꡿쌀 κ³½λŒ€κ·Όμ˜ μ•„λ™μΈκΆŒμΉΌλŸΌ β‘‘] ν•œκ΅­μ˜ ν•™κ΅λŠ” μ•„μ΄μ—κ²Œ μ•ˆμ „ν•œκ°€
    2. The Verge β€” OpenAI says its new cybersecurity model is better at not hacking other companies
    3. OpenAI β€” GPT‑6 Cyber: Advancing the frontier of aligned cybersecurity agents
    4. The Hacker News β€” OpenAI AI Pipeline Generates 948 Linux Root Escalation Exploits in 30-Hour Run
    5. Axios β€” OpenAI's new cyber model found ways to attack nearly 1000 Linux flaws
    6. Fortune β€” OpenAI debuts new cybersecurity-focused AI model amid growing alarm about AI-driven cyber attacks
    7. The Hacker News β€” OpenAI Launches GPT-6 Cyber After AI Helps Find 81 Firefox Flaws, 948 Linux Exploits
    8. WSJ β€” OpenAI Bolsters Cyber Defenses
    9. The Register β€” OpenAI releases GPT-6 Cyber, and promises this one stays within 'rules of engagement'
    10. tech.yahoo.com β€” OpenAI's New GPT-6 Cyber AI Can Find and Exploit Software Flaws β€” It Already Uncovered 588 Unknown Linux Vulnerabilities and 81 Firefox Bugs
    11. tech.yahoo.com β€” GPT-6 Cyber is finding security flaws before hackers can ...
    12. www.bleepingcomputer.com β€” OpenAI's new GPT-6 Cyber AI finds 81 Firefox flaws, 948 Linux ...
    confidence 88%
πŸ“Š

Community Sentiment: How do you assess this situation?

Voice your perspective Β· Real-time aggregated sentiment from the Live Feeds community