OpenAI AI uncovers 948 Linux kernel privilege escalation vulnerabilities
Original title: OpenAI Cyber Model Uncovers 948 Linux Kernel Privilege Escalation Vulnerabilities
AI agents validated 948 Linux kernel flaws in 30 hours, with patches starting within 10 days and bounties valued at $18.9M-$66.3M.
OpenAI's cyber model pipeline scanned Linux kernel source, selecting audit targets covering 2.8M lines of code. Analyst agents generated candidates, triage agents built PoCs, and exploit agents confirmed 1,169 privilege escalations after variant analysis. Maintainers patched 61 upstream in 10 days. The run cost over $1.2M and used new techniques not seen by kernel experts. Real-world chaining of kernel flaws with PHP2Shell has occurred.
- 948 unique vulnerabilities found and patched
- Bounties could reach $66.3M
- New exploitation techniques developed
- Patches sent via Patch the Planet program
Impact: gain root as normal local user
| Product | Affected |
|---|---|
| Linux kernel | — |
| OpenAI cyber model | — |
- 1.Install kernel updates immediately
- 2.Limit unprivileged user namespaces
- 3.Use sandboxes such as gVisor
- Vendors
- OpenAI
- Products
- Linux kernel
- Organizations
- Patch the Planet
From the original article
3 min read at cybersecuritynews.com
OpenAI says an automated research pipeline found and proved 948 separate Linux kernel privilege escalation flaws in one run. These bugs could let a normal local user gain root access. The run took place in mid-October 2026 and used an upcoming cyber model, as OpenAI explains in its research write-up.
By Guru Baran · cybersecuritynews.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.