Snowflake has introduced Cortex AI Gateway, a new platform aimed at providing enterprises with centralized control over how AI agents interact with their systems. The gateway addresses growing concerns about managing multiple AI agents from different platforms that need access to corporate data, applications, and tools.
The platform tackles three primary obstacles organizations face when deploying AI systems: breaking down data silos that prevent AI access, preparing data to meet AI readiness standards, and establishing methods to measure AI data quality. These challenges have become more pressing as enterprises adopt multiple AI agents and models across their operations.
Cortex AI Gateway provides technical controls through support for more than 100 Model Context Protocol (MCP) servers, which serve as standardized interfaces for agent interactions. The platform establishes consistency in how agents access models, data sources, applications, and various tools within the enterprise environment. This standardization helps prevent fragmented approaches to AI agent management.
A key feature of the gateway is comprehensive activity recording that captures agent actions from start to finish across all connected systems. This logging capability allows security and operations teams to analyze what actions AI agents take, providing visibility into automated decisions and system interactions that might otherwise go unmonitored.
According to Mayank Upadhyay, Snowflake's Chief Security and Trust Officer, the platform addresses a fundamental requirement for multi-agent environments. Organizations need assurance about how agents from different platforms access sensitive data, invoke system tools, and perform actions on behalf of users. The gateway aims to provide that trust layer while enabling interoperability between agents from various sources.
Source: https://cybermagazine.com/news/snowflake-launches-cortex-ai-gateway-to-govern-agentic-ai