OpenAI AI model generates 948 Linux kernel root exploits in single run
Original title: OpenAI’s upcoming AI model built 948 working Linux root exploits in a single run
OpenAI's frontier model produced 948 local privilege escalation exploits from Linux kernel bugs at a cost of $1,213,430.
OpenAI's Cyber team used an upcoming AI model in mid-October 2026 to turn Linux kernel bugs into 948 local privilege escalation exploits. The pipeline cost $1,213,430. 2,308 candidate findings were triaged to 1,724 reproducible bugs, leading to 1,169 working root exploits grouped into 948 distinct vulnerabilities. A sample of 166 exploits used 10 new exploitation techniques unknown to experts. The scan took 10 days, with 61 bugs patched upstream. OpenAI recommends keeping Linux systems updated.
Unconfirmed: Details of exploitation techniques kept secret for safety reasons.
- 948 distinct vulnerabilities from 1,169 proofs of concept
- 10 new exploitation techniques discovered by AI agents
- Pipeline ran against latest Linux kernel on stock desktop OS
- 61 bugs patched upstream after 10-day analysis
Impact: local privilege escalation to root access
- Vendors
- OpenAI
- Products
- Linux kernel
- AI models
- frontier model
From the original article
4 min read at helpnetsecurity.com
An automated system running on an upcoming AI model turned hundreds of Linux kernel bugs into working exploits that give an ordinary user root access, and wrote a patch for each underlying flaw, OpenAI revealed in a new research post.
By Sinisa Markovic · helpnetsecurity.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.