Operant AI unveils MCP Gateway to secure emerging runtime AI workflows

Operant AI, the cybersecurity firm behind the AI Gatekeeper platform, has launched MCP Gateway, a significant upgrade aimed at protecting artificial intelligence systems built on the rapidly spreading Model Context Protocol (MCP).

MCP Gateway is positioned as the first enterprise-grade solution providing real-time, active-blocking security for AI workflows that leverage open-source MCP servers and agents. This development arrives amid a surge in MCP adoption—more than 1,000 servers are currently deployed globally—and rising cybersecurity concerns around AI tool integration.

The Model Context Protocol has emerged as a key infrastructure layer for enabling large language models (LLMs) to interact with APIs, databases, and software tools securely. According to Gartner, by 2026, up to 75% of API gateway vendors and half of iPaaS providers are expected to incorporate MCP features. In India alone, where the AI market is forecast to reach \$17 billion by 2027 per NASSCOM, adoption is growing across IT services, financial technology, and healthcare.

Yet, the proliferation of MCP-aligned agents and tools has introduced complex new threats. These include prompt injection attacks, data leakage, and shadow AI deployments. A particularly worrying vector is inter-agent spoofing—where one compromised AI agent may mislead others across interconnected systems. Operant AI argues that conventional cybersecurity tools, designed for static environments, are ill-equipped to handle these dynamic, modular AI architectures.

“While MCP is a powerful AI integration framework, the growing adoption of open-source MCP servers also brings new challenges around security and governance,” said Vrajesh Bhavsar, CEO and co-founder of Operant AI. “Many organizations are deploying these tools without full visibility into their behavior, data access, or usage. MCP Gateway helps close this gap by providing enhanced security, observability, and control across the MCP ecosystem.”

The product offers a triad of capabilities:

1. MCP Discovery enables live mapping of agent behavior and network telemetry across environments—from local development tools like GitHub Copilot to multi-cloud setups.
2. MCP Detections supports advanced risk monitoring for AI-specific threats including jailbreaks, unauthorized access, and supply chain vulnerabilities.
3. MCP Defense brings real-time enforcement, redaction, access control, and encryption to MCP workflows—effectively acting as a runtime firewall for LLMs and agents.

Operant AI is also rolling out a partner program to integrate MCP Gateway with existing AI infrastructure vendors. This co-development model is meant to embed runtime security capabilities directly into partner offerings.

Ashley Roof, co-founder and CMO, said the company’s approach draws on its earlier go-to-market strategies with model providers: “We’re not just securing MCP applications—we’re enabling the entire ecosystem.”

MCP Gateway is available as part of Operant’s broader AI Gatekeeper suite and supports deployment across major cloud platforms, including AWS Bedrock, Azure, and Google Vertex AI.

Founded by veterans from Apple, VMware, and Google, Operant AI is backed by Felicis Ventures and SineWave. The firm has been recognized by Gartner in its Market Guide for AI Trust, Risk, and Security Management (AI TRiSM) as a representative vendor.