AI
OpenAI's agent pipeline finds 948 Linux privilege-escalation flaws for $1.2 million
OpenAI says an autonomous discovery, exploitation and patching pipeline audited about seven percent of the Linux kernel, 5,034 targets and roughly 2.8 million lines, and validated 948 distinct local privilege-escalation flaws with candidate patches. The mid-October run reproduced 1,724 of 2,308 candidate findings and cost $1,213,430 in API credits, about $1,279 per flaw. Reports have gone to maintainers; how many patches land upstream is a separate and still unpublished count.