Anthropic’s Not-So-Secret Weapon That’s Giving Agents a Boost
Every few months, a new acronym takes hold among artificial intelligence researchers.
The latest one gaining steam is MCP, or model context protocol, an open source software Anthropic released in November. It’s a software standard that, if widely adopted, would help large language models of any kind access external data sources and tools, such as a website or database. Developers have described MCP as a kind of “universal translator” for models to be able to talk to the web or applications.
In the past, developers have had to write custom code so that an AI model can do a web search or pull information from a database, for instance.
But now, MCP offers a standardized way for AI to connect with external tools and services to look up information as well as take actions, such as searching your Gmail or posting a message to Slack.
Related Knowledge Center: Artificial intelligence
On top of a failing business model, newsrooms must navigate the powerful — and not fully reliable — new capabilities of artificial intelligence and the deployment of large language models.
State of the business model | Copyright and intellectual property