OpenAIs Cyber-KI findet Dutzende Sicherheitslücken bei Apple

29 Quellen
  • OpenAI und die Sicherheitsfirma Calif nutzten GPT-6 Cyber, um etwa 35 Schwachstellen in Apple -Produkten zu finden, darunter eine Umgehung des Lockdown-Modus, berichtete Forbes am Dienstag.
  • Das Modell debütierte am 5. November, nachdem OpenAI bekannt gab, dass es autonom 948 Exploits für den Linux-Kernel zu Kosten von etwa 1,2 Millionen Dollar erstellt hatte.
  • Apples Sicherheitshinweise schreiben OpenAI bisher nur drei WebKit-Korrekturen zu, während Calif-CEO Thai Duong vor "mehr Fehlern, mehr Exploits" in der Zukunft warnt.
Quellen (29)
  1. 1 OpenAI's hacking model uncovers dozens of security flaws in Apple devices www.forbes.com
  2. 2 A thousand fewer paths to root openai.com
  3. 3 Claim that OpenAI model found 948 Linux flaws runs ahead of the record cryptobriefing.com
  4. 4 AI Agents Got Root on Linux via 948 Kernel Flaws in OpenAI Research www.securityweek.com
  5. 5 OpenAI says it found 948 local privilege escalation ... www.theverge.com
  6. 6 OpenAI Daybreak: NATO's NCIA runs defender models at up to 300 billion tokens a month, and allies fix AI-found vulnerabilities · IA Compa iacompa.com
  7. 7 OpenAI AI Builds 948 Linux Kernel Root Exploits for $1.2M - Cyber Kendra www.cyberkendra.com
  8. 8 OpenAI used AI to find nearly 1,000 ways for hackers to hijack Linux www.howtogeek.com
  9. 9 November 18, 2026 reg4tech.com
  10. 10 Who protects the antivirus? How we secure Malwarebytes itself www.malwarebytes.com
  11. 11 OpenAI cyber model found 948 Linux kernel exploits www.artificialintelligence-news.com
  12. 12 OpenAI's agents found 948 Linux root bugs for $1.2 million | ANOTHER News anothernews.io
  13. 13 AI University Now Urges Executives to Track How Long Security Holes Stay Unpatched as AI Speeds Discovery www.elpasotimes.com
  14. 14 Java4Shell (CVE-2026-114938): Affected Versions, Fix, Workaround www.cyberkendra.com
  15. 15 AI Finds Zero-Day Flaws: 66 in 90 Hours for $935 techreview.si
  16. 16 OpenAI Finds Three Apple Flaws and 948 Linux Vulnerabilities demokratis.co.id
  17. 17 OpenAI Launches GPT-6 Cyber That Finds 81 Firefox Flaws and … cybersecuritynews.com
  18. 18 OpenAI Codex Security Credited in Apple Patches | Wisevoter wisevoter.com
  19. 19 OpenAI Reports 948 Linux Kernel Root Exploits from an Automated Pipeline, Then Releases the Model as GPT-6 Cyber — The Ledger www.the-ledger.net
  20. 20 systemcards.org www.systemcards.org
  21. 21 OpenAI's 948 Linux Kernel Bugs Are Only the First Test dev.to
  22. 22 OpenAI says its cyber model patched 948 Linux kernel... daily.dev
  23. 23 GPT-6 Cyber Opens to Vetted Users: An Exploit Still Needs to Prove ... nilsliu.dev
  24. 24 Research openai.com
  25. 25 OpenAI's 948 Linux Root Exploits: What to Patch | Beam agentbeam.com
  26. 26 OpenAI's GPT-6 Cyber found about 35 flaws in Apple's iPhone and Mac software, Forbes reports mlq.ai
  27. 27 OpenAI Daybreak: la OTAN usa modelos defensores con hasta 300.000 millones de tokens al mes · IA Compa iacompa.com
  28. 28 OpenAI frontier model breaks @Apple but continues to expose vulnerabilities ... x.com
  29. 29 OpenAI Reveals AI's Role in NATO Cyber Defense forklog.com