These protocols will help AI agents navigate our messy lives
1 min read
Summary
Google and Anthropic, along with other groups, are developing protocols for artificial intelligence (AI) agents in an effort to make them more effective at completing tasks.
This is due to the inconsistent nature of AI systems, which means application programming interfaces are unable to communicate successfully with these randomised systems.
Google has introduced the Agent2Agent protocol, which looks at how AI agents interact and how they can communicate effectively, whilst Anthropic’s Model Context Protocol (MCP) focuses on natural language and how agents can interact with various programs and applications.
However, both protocols have three main areas of growth – security, openness and efficiency – which still require work according to analysts.
Security is particularly important because of the increasing number of attacks on AI systems, whilst the other two growth areas could be crucial in making these protocols as effective as possible going forward.