OpenAI AI pipeline finds exploits and patches Linux kernel flaws
Original title: OpenAI built an AI pipeline to find, exploit and patch Linux flaws
OpenAI's pipeline identified 948 Linux local privilege-escalation vulnerabilities at $1.2 million API cost through automated discovery, exploitation and patching.
OpenAI developed an autonomous AI pipeline that identifies software vulnerabilities, develops exploits and generates patches for the Linux kernel. In an October experiment, it created exploits for 948 distinct LPE vulnerabilities at a total cost of $1.2 million. The system automates discovery, validation and patch generation, with human researchers reviewing outputs before reporting to maintainers. Some vulnerabilities were already patched upstream by the time of reporting.
Unconfirmed: Not all 948 findings were submitted for reporting; some patches introduced unnecessary code changes
- Pipeline created 948 LPE exploits at $1,279 average cost each
- Identified five times more vulnerabilities than prior GPT experiment despite kernel hardening
- At least 10 novel Linux exploitation techniques used
- 61 vulnerabilities already patched upstream
- Hundreds of reports and patches submitted to maintainers
Impact: AI-generated exploits for local privilege escalation in Linux kernel
- 1.Submit AI-generated reports and patches to maintainers
- 2.Apply upstream fixes when available
- 3.Validate AI exploits independently
- Vendors
- OpenAI
- Products
- Linux kernel
- Organizations
- Linux kernel maintainersSecurity BrigadeIDCSUSE
From the original article
3 min read at csoonline.com
OpenAI has developed an autonomous AI pipeline that can identify software vulnerabilities, develop working exploits to validate their impact and generate patches for human review. In an October experiment targeting the Linux kernel, the pipeline created exploits for 948 distinct local privilege-escalation (LPE) vulnerabilities at a total API cost of $1.
csoonline.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.