微软周一推出了包括Guardian Agents在内的一系列新安全功能，旨在防护企业面临的AI代理威胁1。这些工具是Project Perception的组成部分，用于监控企业部署和使用的AI代理1。微软首席执行官纳德拉在私人博客中强调，企业需要采取更谨慎的态度应对AI模型提供商，主张"将智能供应与控制权分离"1。
微软安全业务主管Hayete Gallot指出，前沿模型因其不可预测性而无法信任，无论其来源如何1。为此，微软在Microsoft Defender中添加了代理检测功能，可识别60种类型的AI代理1。此外，微软已成为Nvidia开放代理安全平台的创始成员之一，该平台已获得100多家科技公司的支持1。
Microsoft introduced Guardian Agents and other security features on Monday to protect enterprises from threats posed by artificial intelligence agents.1 The company's perspective reflects growing concerns about the reliability and controllability of advanced AI systems in business environments.
CEO Satya Nadella emphasized in a private blog post that "we need to decouple the supply of intelligence from the control of it," underscoring the necessity for organizations to maintain independent oversight of AI deployments.1 Hayete Gallot, head of Microsoft's security business, stated: "Unfortunately, frontier models cannot be trusted regardless of their source," citing their inherent unpredictability as a fundamental challenge.1
Guardian Agents, developed as part of Project Perception, will monitor AI agents deployed within enterprise environments.1 Microsoft has also integrated agent detection capabilities into Microsoft Defender, enabling identification of 60 different types of agents.1 Additionally, Microsoft joined Nvidia's open agent security platform as a founding member, which has already attracted support from over 100 technology companies.1
评论
还没有评论，欢迎留下第一条。