AI-related risk source: Risk can arise at every stage of an AI system's life, from inception and design through development, verification, deployment, operation, monitoring and retirement, and from transitions between stages, including changes and updates.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.