Big TechTechCrunch

Microsoft says AI agents are like insider threats as ‘frontier models can’t be trusted’

#microsoft#ai#security#enterprise#insider threats

English

Microsoft has introduced new security features, including Guardian Agents, to protect enterprises from potential threats posed by AI agents, which CEO Satya Nadella likens to insider threats. He emphasizes the need for enterprises to monitor these agents independently, as they cannot be fully trusted due to their unpredictability and access to sensitive data.

中文

微软推出了新的安全功能，包括Guardian Agents，以保护企业免受AI代理可能带来的威胁。首席执行官萨提亚·纳德拉将这些代理比作内部威胁，强调企业需要独立监控这些代理，因为它们由于不可预测性和对敏感数据的访问而不能完全信任。