Event ·Multi-source

OpenAI 发布 GPT-6 Cyber,漏洞挖掘能力超 GPT-5.6 Sol

OpenAI 发布安全模型 GPT-6 Cyber,官方称其表现超过 Astra。在一次漏洞研究中,该模型为 Mozilla 确认 81 个高严重性新漏洞并提供可复现测试,同时指出 30 个已知问题比此前评估更严重。针对 Linux 内核,GPT-6 Cyber 复现 1,724 个漏洞并开发出 948 个可用 exploit,是 GPT-5.6 Sol 同样条件下 166 个的五倍以上。OpenAI 同步成立 Daybreak Red for Critical Infrastructure 团队,服务关键基础设施防御者。

Current read

OpenAI 新出的 GPT-6 Cyber 在 Linux 内核上做出了 948 个可用 exploit,是 GPT-5.6 Sol 的五倍多,搞安全的朋友可以看看。

11 sources91° AI heatLast updated 2026/11/5 17:25:52

Evidence

11 sources
Primary sourceAndrew Curran
View source
Related source 2kimmonismus
View source
Related source 4TestingCatalog
View source
Related source 6Greg Brockman
View source
Related source 7rohanpaul_ai
View source
Related source 9Yuchen Jin
View source
Related source 10coderabbitai
View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.