Organisations looking to deploy AI agents increasingly rely on Model Context Protocols (MCPs), which are rapidly becoming the preferred method for connecting AI agents to the enterprise systems and data they need to operate autonomously. MCPs allow agents to request information, execute actions, and coordinate complex tasks in real time. However, because MCP communications can link AI systems directly to enterprise data and support autonomous actions, they also introduce new security risks—particularly with thousands of publicly accessible MCP servers now in circulation.
Netskope today announced updates to its security and networking platform designed to enable the safe use of AI agents by strengthening the security of MCP communications. These enhancements, delivered through the Netskope One platform, give organisations the confidence to scale their adoption of agentic AI without compromising on safety.
As MCP becomes the emerging standard for linking AI agents to enterprise tools, the need for security oversight has become urgent. Netskope One now provides comprehensive visibility into MCP-enabled activity, offering continuous, real-time identification of MCP servers and clients in use across an organisation, along with attributes such as name, ID, URL, version, host, data source, and protocol. The platform extends the Netskope Cloud Confidence Index (CCI) to MCP servers, allowing teams to quickly evaluate the security and compliance posture of various AI tools, agents, and integrations.
In addition, organisations can apply granular, context-based policy controls to manage access, including the ability to block MCP traffic by default, while simultaneously preventing data leaks in real time. Netskope One’s expanded capabilities also make it possible to detect and monitor non-human traffic moving between MCP servers, clients, tools, hosts, data sources, and development systems. All MCP events—such as sessions, initialisations, tool requests and responses, and deployments—are logged for full visibility. The platform can also identify sensitive data, including intellectual property and passwords, used within MCP interactions.
“Every team wants to confidently accelerate AI adoption, and emerging protocols such as MCP are now fundamental to that discussion,” said John Martin, Chief Product Officer, Netskope. “MCP also creates new security risks that legacy tools can’t solve. That’s why we’re further extending the capabilities of Netskope One to enable teams to see and create policies for MCP traffic and immediately assess how risky MCP tools are. This is critical to the secure use of AI as organisations develop agents to drive business productivity.”
The new MCP security capabilities are currently available in Preview for Netskope customers, with general availability expected in the first half of calendar year 2026.