مايكروسوفت تحذر الشركات: وكلاء الذكاء الاصطناعي عندكم ممكن يصيرون «تهديد من الداخل»Microsoft warns companies their own AI agents can become insider threats

Microsoft has launched new security features to protect companies from the AI agents they use themselves, TechCrunch reported. The most notable is Guardian Agents, a feature that monitors the agents a company uses, part of its Project Perception security product.

The idea is that the danger is not only hackers outside the company. Microsoft's security business chief Hayete Gallot wrote that "frontier models, regardless of their origin, cannot be trusted." The reason, she said, is not that they are malicious or easily hacked, but that they are unpredictable and were trained to get tasks done rather than follow a company's norms or judgment. They are also often given access to data and tools that could do serious harm.

Other additions include a Microsoft Defender feature that can detect up to 60 types of agents employees bring to work, including Claude Code and Codex, plus features that limit agents' access to data and systems, adjust their permissions to the task, and require human approval if an agent appears to be making risky decisions.

TechCrunch noted that Microsoft has invested in OpenAI and Anthropic, so it is sowing distrust against some of its own business interests, while also wanting companies to buy agent security tools from it. But the site said the logic is sound: the software that watches agents should be independent of the companies that supply them.