The Model Context Protocol (MCP) is reshaping how AI assistants access real-time information. Launched by Anthropic in November 2024 and now governed by the Agentic AI Foundation, MCP enables AI tools like Claude, ChatGPT, and Cursor to query data directly from publisher servers rather than relying on static training data. This decentralised model ensures answers are current, sourced, and licensed, shifting power from centralised search engines to individual publishers.
By exposing data through standardised servers, organisations—from statistics offices to specialised websites—can provide verified, up-to-date responses to AI queries without intermediaries. The protocol also mandates clear citation, trust boundaries, and privacy-preserving analytics. Contexte Tech, a European open-data platform, has implemented an MCP server offering multilingual datasets, prompts, and tools with built-in attribution and usage tracking, demonstrating how the ecosystem can evolve toward transparent, accountable AI search.
Source: Hacker News · Summarized by HeadlinesBriefing