AI & Agentic Tools
Bug economics
OpenAI says an unreleased model produced 948 Linux root exploits for about $1,279 each
OpenAI reported on November 4 that an unreleased cyber model ran an autonomous pipeline against the Linux kernel in mid-October and produced 948 local privilege-escalation exploits, each with a proof of concept and a candidate patch. Total API spend was $1,213,430, roughly $1,279 per unique flaw. The figures are OpenAI's own, and the model has not been released, so none of it has been independently checked.