Microsoft adds tools to limit and monitor AI agents in Microsoft 365 E7

Microsoft has introduced new tools for Microsoft 365 E7 and Project Perception to monitor and control AI agents. These features assess the risk of each agent, limiting their reach and tightening restrictions as risk levels increase. The system allows companies to block agents from making changes, require human intervention, or shut them down entirely. Hayete Gallot, executive vice president of Microsoft Security, detailed this approach to help organizations effectively manage and corral rogue AI agents.