Microsoft adds Guardian Agents and tighter controls over enterprise AI agents

Microsoft introduced Guardian Agents on Monday to monitor AI agents used by enterprises, TechCrunch reported. The feature joins Project Perception, which previously focused on agents that attack systems, assess vulnerabilities and fix them. Microsoft also added detection to Defender for up to 60 types of agents brought into workplaces by employees, including Claude Code, Codex and OpenClaw. Additional controls restrict agents' access to data and systems, adjust permissions to the task and require human approval when an agent appears to make risky decisions. Microsoft security chief Hayete Gallot said the concern is that models are unpredictable and trained to accomplish tasks rather than follow organizational judgment.