微软扩展 Project Perception，新增智能体风险持续评估与权限自适应能力

Satya Nadella· @satyanadella · X·2026-10-19 21:05· 1 小时前AI 评分41

AI 导读

我最近撰文谈到，为什么我们需要把模型视为内部风险，并将安全控制置于模型本身之外。 今天，我们正把这一原则付诸实践：扩展 Project Perception，新增能力以帮助持续评估智能体风险、自适应调整权限并遏制威胁。 https://blogs.microsoft.com/blog/2026/10/19/securing-the-agentic-insider/

正文 · 原文

I recently wrote about why we need to treat models as insider risks, with security controls sitting outside the model itself.

Today, we're putting that principle into practice by expanding Project Perception with new capabilities to help continuously assess agent risk, adapt permissions, and contain threats.
https://blogs.microsoft.com/blog/2026/10/19/securing-the-agentic-insider/

来源：Satya Nadella · x.com