Microsoft introduced the Guardian Agents feature to oversee corporate AI agents
Microsoft introduced the Guardian Agents feature in Project Perception to monitor corporate AI agents. It cites the unpredictability of models as the reason oversight is needed and recommends separating the provision of models from the authority to control their actions.
Microsoft introduced the Guardian Agents feature, which is intended to monitor AI agents used by companies. The feature is part of Project Perception, a product introduced in June. It expands the product’s focus from external attackers to include risks associated with agents the company itself has deployed.
According to Microsoft, advanced models are unpredictable, and training them to perform tasks does not guarantee compliance with company rules. Microsoft chief Satya Nadella therefore recommends establishing separate systems for overseeing and managing agents that are not controlled by model providers. He warns that a provider’s assurances about safety do not absolve a company of responsibility for the actions of its agents.
Why it matters
Security teams are expected to gain a tool for monitoring agents that the company authorizes to access sensitive data and perform important tasks. This also makes oversight of the behavior of already approved agents as they work a practical concern.
Relevant practical impact
What this means
For a business
Deploying AI agents raises the question of who within a company oversees their actions and bears responsibility for them. Microsoft recommends separating this oversight from the model provider.
Risks and complianceCheck the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.