⚡Megatron LLM Hub
←返回搜索
Twitter

OpenAI 用自主流水线一次验证并修补 948 个内核本地提权漏洞

显示自动化漏洞挖掘与修补的成本产出比,Agent 攻防风向标。
ai_agentlpe内核自动化
RT @OpenAI: Our security researchers built an autonomous cyber vulnerability discovery, exploitation, and patching pipeline. In a single run with an upcoming cyber model, it validated and generated patches for 948 local privilege escalation vulnerabilities in the Linux kernel at a total API cost of $1.2 million.