Home / Wiki / M / Model Context Protocol (MCP)

Model Context Protocol (MCP)

Technology and VendorsUpdated 16 Nov 2026

The Model Context Protocol (MCP) is an open standard that defines how an application built on a large language model connects to outside data sources and software tools. It was introduced by the AI company Anthropic in November 2024 and published as open source. In December 2025 Anthropic gave the protocol to the Agentic AI Foundation, a fund set up under the Linux Foundation. A program called an MCP server describes what it offers, such as files to read or actions to perform, in a common format. Any compatible AI application, acting as client, can then discover and call those functions without a connector written specially for it.

In finance, MCP is one of the building blocks of agentic AI. Data vendors, brokers and payment companies have published servers that let an AI agent query prices, read account records or start a transaction, which supports agentic payments. The main concerns are security and control: a model can be misled by malicious instructions hidden in the data it reads, so permissions and human approval matter.

Last updated 16 Nov 2026