Citrix NetScaler MCP Gateway Adds Unified Security for Agentic AI Traffic

Citrix, a Cloud Software Group company, has introduced Model Context Protocol (MCP) Gateway capabilities for NetScaler, extending its application delivery and security platform to govern enterprise agentic AI traffic. Announced on July 9, 2026, the update allows organizations to centrally route, authenticate, monitor, and control AI agent requests to backend MCP servers. Citrix also expanded […] The post Citrix NetScaler MCP Gateway Adds Unified Security for Agentic AI Traffic appeared first on Cyber Security News.

Citrix, a Cloud Software Group company, has introduced Model Context Protocol (MCP) Gateway capabilities for NetScaler, extending its application delivery and security platform to govern enterprise agentic AI traffic. Announced on July 9, 2026, the update allows organizations to centrally route, authenticate, monitor, and control AI agent requests to backend MCP servers. Citrix also expanded NetScaler AI Gateway with policy-based large language model (LLM) routing and token-level usage visibility, positioning the platform as a unified control point for both agent and model traffic. Citrix NetScaler MCP Gateway MCP is rapidly emerging as a standard mechanism for connecting AI agents with enterprise tools, data stores, and systems of record. However, organizations deploying MCP infrastructure can face the same governance issues historically associated with unmanaged APIs: fragmented endpoints, inconsistent authentication, insufficient logging, and uncontrolled request volume. NetScaler MCP Gateway is designed to address those concerns by providing a single governed ingress point for MCP clients. Rather than requiring teams to configure connections and identity controls for each MCP server, enterprises can dynamically route requests to approved backend services via centrally managed policies. The gateway supports per-user and global token controls, as well as OAuth and hybrid authentication flows. Security teams can also configure server allow and block lists, as well as tool-level rate limits intended to restrict unapproved agent actions and reduce the risk of runaway automated requests. Citrix said the capability is particularly relevant to regulated sectors, including financial services, healthcare and government, where AI agents may access sensitive applications and information. Centralized authentication, policy enforcement and auditability can help organizations retain control as agents execute multi-step workflows across internal environments. NetScaler MCP Gateway also includes session persistence and protocol-aware health monitoring. These functions are intended to keep an agent connected to the appropriate backend MCP server throughout longer workflows and to detect unavailable or unhealthy services. Alongside the MCP Gateway release, Citrix expanded NetScaler AI Gateway to support content-switching-based model routing. Enterprises can use routing policies to send incoming AI application and agent requests to different LLM providers or models based on operational requirements. The platform can additionally track input and output tokens, as well as request activity, by user, team or application. This visibility may help organizations allocate AI costs, monitor consumption patterns, and avoid dependency on a single model provider. Citrix said the features enable administrators to apply centralized governance across both the model layer and the agent-to-tool layer. The company is also privately previewing a Claude Code use case. In this deployment, NetScaler AI Gateway operates in front of Anthropic model access, serving as a centralized authentication and routing layer for developers. Citrix emphasized NetScaler’s single-pass architecture as a key element of the platform’s AI governance approach. The architecture performs authentication, traffic management, security inspection, routing, rate limiting, and observability in a single data-path pass. This design aims to reduce additional hops, latency, and CPU overhead when processing high-volume AI workloads. Citrix said it already uses NetScaler AI Gateway to control prompts, model interactions, and token usage for its internal Citrix Aidrien AI assistant. As enterprises move agentic AI projects from proof-of-concept deployments into production, centralized MCP governance could become an essential security requirement for controlling access to business-critical systems. Follow us on Google News , LinkedIn and X to Get More Instant Updates. Set Cyberpress as a Preferred Source in Google. The post Citrix NetScaler MCP Gateway Adds Unified Security for Agentic AI Traffic appeared first on Cyber Security News.