Citrix has announced NetScaler MCP Gateway capabilities designed to route, govern, and observe traffic between AI agents and backend MCP servers, alongside NetScaler AI Gateway updates for model routing and token-level usage tracking.
Citrix announced new NetScaler MCP Gateway capabilities to help enterprises route, govern, and observe traffic between AI agents and backend Model Context Protocol servers.
Citrix said its NetScaler MCP Gateway functionality is intended to provide a governed entry point for agent traffic moving between AI clients and backend MCP servers. The company described the update as part of a broader effort to bring enterprise controls to large language model and agentic AI traffic.
Help Net Security separately reported that Citrix added NetScaler MCP Gateway to secure enterprise AI agents, corroborating Citrix’s description of the feature as a way to manage agent-to-MCP-server traffic.
The Model Context Protocol, or MCP, is used to connect AI systems with tools, applications, and data sources. IT Brief Asia reported that Citrix’s NetScaler update gives enterprises a single AI traffic entry point as agents connect to internal applications, tools, and data through MCP servers.
NetScaler documentation describes the product operating as a centralized MCP Gateway. According to the documentation, NetScaler can use content switching and load balancing to route MCP clients from a single virtual endpoint to backend MCP servers.
That architecture is aimed at centralizing traffic management rather than requiring each AI agent or MCP server connection to be handled independently. Citrix’s announcement said the MCP Gateway capabilities are designed to route, govern, and observe agent traffic to backend MCP servers.
For enterprises experimenting with multiple AI agents and internal data connections, the gateway model may offer a more familiar operational pattern: traffic is directed through a controlled intermediary where routing and visibility can be applied.
Citrix also announced upgrades to NetScaler AI Gateway. According to the Citrix announcement, those updates include model routing and token-level usage tracking for LLM traffic.
Help Net Security also reported the expanded AI Gateway capabilities, including model routing and token-level LLM usage tracking. Those functions can help organizations see how AI services are being used and direct requests across models according to policy or operational needs.
The source materials do not provide detailed pricing or deployment timelines beyond the announced NetScaler capabilities. They do, however, position the MCP Gateway and AI Gateway updates as part of Citrix’s effort to apply existing traffic governance patterns to newer AI workloads.
As companies connect AI agents to internal systems, the traffic between agents, tools, data sources, and models can become harder to monitor. Citrix’s announcement frames NetScaler MCP Gateway as a way to bring centralized routing, governance, and observability to that environment.
The practical significance is that enterprises may be able to manage AI agent connections through infrastructure they already associate with application delivery and traffic control. NetScaler documentation supports that framing by describing the gateway as a central virtual endpoint that routes MCP clients to backend MCP servers using established NetScaler functions such as content switching and load balancing.
Citrix is not claiming that the gateway alone solves all AI security or governance issues. Based on the company’s announcement and corroborating coverage from Help Net Security and IT Brief Asia, the update is best understood as an infrastructure control point for managing how AI agents and LLM-related services communicate across enterprise environments.
Citrix announced new NetScaler MCP Gateway capabilities to help enterprises route, govern, and observe traffic between AI agents and backend Model Context Protocol servers.
A gateway layer for agentic AI traffic Citrix said its NetScaler MCP Gateway functionality is intended to provide a governed entry point for agent traffic moving between AI clients and backend MCP servers.
The company described the update as part of a broader effort to bring enterprise controls to large language model and agentic AI traffic.
Continue reading
