Anthropic has launched a new open-source tool aimed at establishing direct connections between AI assistants and the necessary information to support their responses or task executions. This tool, named the Model Context Protocol (MCP), offers a universal interface that can connect to a wide range of data sources, which Anthropic claims will enhance AI performance.

Earlier this month, OpenAI began testing a feature called "Collaborating with Applications," allowing the Mac version of ChatGPT to directly connect to certain programming applications. In comparison, Anthropic's tool is designed to be compatible with all AI systems and data sources.

Alex Albert, Head of Claude Relations at Anthropic, pointed out that currently, developers need to write custom code for each dataset they want the AI model to access. With Anthropic's MCP, developers only need to integrate MCP with their AI toolkit once, enabling connections to any data source through a "standard protocol for sharing resources, tools, and prompts."

Anthropic disclosed that programming software such as Replit, Codeium, and Sourgraph have started using MCP to build their AI agents, which can perform tasks on behalf of users. This tool could help other companies and developers link AI systems to multiple data sources, which is particularly important as the AI industry moves towards agent-based AI.

In their announcement, Anthropic stated, "Developers no longer need to maintain separate connectors for each data source; instead, they can develop based on a standard protocol. As the ecosystem matures, AI systems will retain context while moving between different tools and datasets, replacing the current fragmented integrations with a more sustainable architecture."