Reports

IDC Innovators: Security for Agentic AI, 2025

This IDC Innovators presentation highlights emerging vendors addressing the security challenges of agentic AI, particularly around the Model Context Protocol (MCP) — a rapidly adopted standard for AI agent communication. And no, MCP isn’t a new Marvel villain — though it might be just as disruptive.

While MCP enables seamless integration across tools and systems, it lacks robust security specifications, creating vulnerabilities in enterprise environments. The content for this excerpt was taken directly from the document “IDC Innovators: Security for Agentic AI, 2025”, September 2025