Netskope Expands Security Controls for AI Agents Using Model Context Protocol
SANTA CLARA, Calif. — Netskope is advancing protections for organizations adopting agentic AI with new security capabilities for Model Context Protocol (MCP). This emerging standard allows AI agents to connect to enterprise systems, data sources, and tools.
The company announced that these innovations are now available in preview as part of its Netskope One platform, giving security and IT teams increased oversight and control over MCP communications, enabling AI models to execute tasks and access information autonomously.
MCP is gaining traction as a preferred integration method for AI agents, but its rapid adoption has also introduced new risks. With thousands of publicly available MCP servers and limited transparency into how they are used, organizations face challenges in managing exposure, enforcing access, and preventing sensitive data from being misused.
Netskope’s new capabilities are designed to mitigate those risks by giving enterprises granular insight into MCP activity and the tools behind it. With these updates, better visibility, control, and compliance enforcement can be applied across AI-agent interactions.
Through Netskope One, organizations can now:
• Continuously identify MCP servers and clients in use, along with metadata such as name, ID, URL, version, host, data source, and protocol
• Apply Netskope Cloud Confidence Index (CCI) risk scoring to MCP servers to prioritize risk across AI tools, agents, and integrations
• Enforce context-aware access policies, including the option to block MCP traffic by default
• Detect and monitor non-human traffic between MCP servers, clients, tools, and data sources
• Log key MCP activity, including sessions, initialization, tool requests and responses, and deployments
• Identify sensitive data — including intellectual property and credentials — used within MCP workflows
“Every team wants to accelerate AI adoption confidently, and emerging protocols such as MCP are now fundamental to that discussion,” said John Martin, chief product officer at Netskope. “MCP also creates new security risks that legacy tools can’t solve. By extending the Netskope One platform to identify and apply policy to MCP traffic, teams can immediately understand and reduce risk as they develop and deploy AI agents.”
Netskope expects the new MCP security capabilities to reach general availability in the first half of calendar year 2026.
The company will demonstrate the enhanced Netskope One platform this week at the AWS event in Las Vegas. Attendees can engage with Netskope at booth #460, attend scheduled speaking sessions, and meet with experts for live demonstrations.
About Netskope
Netskope (NASDAQ: NTSK) provides modern security and networking solutions for the cloud and AI era. Its Netskope One platform enables optimized access and real-time, context-based security for users, devices, and data across cloud, AI, SaaS, web, and private applications. More than 30% of the Fortune 100 and thousands of customers globally rely on Netskope’s Zero Trust Engine and NewEdge network to improve visibility, reduce risk, and accelerate performance without compromise.