OpenAI's AI pipeline built 948 working Linux root exploits in a 30-hour run
OpenAI's AI pipeline built 948 working Linux root exploits in a 30-hour run
OpenAI's AI pipeline built 948 working Linux root exploits in a 30-hour run
Key points
اہم نکات
मुख्य बातें
- OpenAI said on 4 November that an automated security pipeline built on an upcoming cyber-focused AI model produced 948 distinct Linux local privilege-escalation exploits in a 30-hour run in mid-October.
- The pipeline examined 2,308 candidate kernel findings, reproduced 1,724 vulnerabilities inside isolated virtual machines, and turned them into working exploits that could let an ordinary local user gain root access, along with proofs of concept for 291 information leaks.
- The run cost about 1.21 million dollars in API credits, around 1,279 dollars per exploit, and the researchers said the agents used at least 10 exploitation techniques they considered novel.
- OpenAI said on 4 November that an automated security pipeline built on an upcoming cyber-focused AI model produced 948 distinct Linux local privilege-escalation exploits in a 30-hour run in mid-October.
- The pipeline examined 2,308 candidate kernel findings, reproduced 1,724 vulnerabilities inside isolated virtual machines, and turned them into working exploits that could let an ordinary local user gain root access, along with proofs of concept for 291 information leaks.
- The run cost about 1.21 million dollars in API credits, around 1,279 dollars per exploit, and the researchers said the agents used at least 10 exploitation techniques they considered novel.
- OpenAI said on 4 November that an automated security pipeline built on an upcoming cyber-focused AI model produced 948 distinct Linux local privilege-escalation exploits in a 30-hour run in mid-October.
- The pipeline examined 2,308 candidate kernel findings, reproduced 1,724 vulnerabilities inside isolated virtual machines, and turned them into working exploits that could let an ordinary local user gain root access, along with proofs of concept for 291 information leaks.
- The run cost about 1.21 million dollars in API credits, around 1,279 dollars per exploit, and the researchers said the agents used at least 10 exploitation techniques they considered novel.
OpenAI said on 4 November that an automated security pipeline built on an upcoming cyber-focused AI model produced 948 distinct Linux local privilege-escalation exploits in a 30-hour run in mid-October. The pipeline examined 2,308 candidate kernel findings, reproduced 1,724 vulnerabilities inside isolated virtual machines, and turned them into working exploits that could let an ordinary local user gain root access, along with proofs of concept for 291 information leaks. The run cost about 1.21 million dollars in API credits, around 1,279 dollars per exploit, and the researchers said the agents used at least 10 exploitation techniques they considered novel. OpenAI said it is reporting the flaws to Linux maintainers through coordinated disclosure and has drafted candidate patches, stressing that the 948 figure counts distinct flaws, not patches already merged into the kernel.
OpenAI said on 4 November that an automated security pipeline built on an upcoming cyber-focused AI model produced 948 distinct Linux local privilege-escalation exploits in a 30-hour run in mid-October. The pipeline examined 2,308 candidate kernel findings, reproduced 1,724 vulnerabilities inside isolated virtual machines, and turned them into working exploits that could let an ordinary local user gain root access, along with proofs of concept for 291 information leaks. The run cost about 1.21 million dollars in API credits, around 1,279 dollars per exploit, and the researchers said the agents used at least 10 exploitation techniques they considered novel. OpenAI said it is reporting the flaws to Linux maintainers through coordinated disclosure and has drafted candidate patches, stressing that the 948 figure counts distinct flaws, not patches already merged into the kernel.
OpenAI said on 4 November that an automated security pipeline built on an upcoming cyber-focused AI model produced 948 distinct Linux local privilege-escalation exploits in a 30-hour run in mid-October. The pipeline examined 2,308 candidate kernel findings, reproduced 1,724 vulnerabilities inside isolated virtual machines, and turned them into working exploits that could let an ordinary local user gain root access, along with proofs of concept for 291 information leaks. The run cost about 1.21 million dollars in API credits, around 1,279 dollars per exploit, and the researchers said the agents used at least 10 exploitation techniques they considered novel. OpenAI said it is reporting the flaws to Linux maintainers through coordinated disclosure and has drafted candidate patches, stressing that the 948 figure counts distinct flaws, not patches already merged into the kernel.
Most read
سب سے زیادہ پڑھی گئی
सबसे ज़्यादा पढ़ी गई
- Loading...
Get news in your inbox
خبریں اپنے ان باکس میں حاصل کریں
खबरें अपने इनबॉक्स में पाएं
One email when new stories are published. No spam, unsubscribe anytime.
نئی خبریں شائع ہونے پر ایک ای میل۔ کوئی سپیم نہیں، کسی بھی وقت ان سبسکرائب کریں۔
नई खबरें प्रकाशित होने पर एक ईमेल। कोई स्पैम नहीं, कभी भी अनसब्सक्राइब करें।