AI-related risk source: Where the system's workings or the reasons for its outputs cannot be understood, errors go undetected, accountability weakens and stakeholders lose trust.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.