Sector players should keep humans in the loop across the four phases of the AI lifecycle (design, data collection, algorithm development and training; verification and validation; deployment; operation and monitoring), so that the data AI generates is managed by people with the expertise to make informed decisions, with user interfaces and interaction mechanisms that support collaboration, transparency and accountability between humans and the system. Stage-specific tools and metrics (diverse perspectives at design, testing and validation tools and metrics, decision thresholds and response times in operation) support each phase.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.