Trend Micro Outlines Four-Part Defense for Agentic AI Risks

The rapid adoption of large language models and agentic AI is expanding the corporate attack surface as autonomous agents gain the ability to access data, invoke software tools and execute tasks. Trend Micro says conventional security controls centered on users and endpoints may not adequately address systems whose permissions and behavior can change dynamically, making clearer accountability and oversight essential to managing the next generation of AI-related risk.

Trend Micro, through its TrendAI initiative, has outlined a four-part framework focused on visibility, ownership, observability and governance. The company recommends that enterprises inventory AI agents and connected tools, assign responsibility for their operation, continuously monitor permissions and behavior, and enforce consistent policies across deployments. The report did not disclose a publication date, spending commitment, customer adoption figures or a timetable for implementing the framework.