Artificial intelligence / news brief
OpenAI reports 948 Linux privilege-escalation findings
- Event
- Published

OpenAI says an autonomous pipeline scanned about 7% of the Linux kernel, reproduced 1,724 vulnerabilities and reduced 1,169 working local privilege-escalation exploits to 948 distinct findings with candidate patches. The mid-October run cost $1.21 million in API credits; most fixes had not yet been merged when OpenAI published its report.
Why it matters
Automated vulnerability research is reaching unusual scale, but exploit attribution, patch quality and dual-use risks still require human review.
Sources
- A thousand fewer paths to root (external site)OpenAI · primary source ·
- OpenAI’s cyber agents find 948 Linux flaws, uncover 10 ways to break system defenses (external site)Interesting Engineering · coverage ·