An analysis of how functions are allocated between human and AI agents, and of how the AI behaves, in every functional mode and at every autonomy level, is conducted to evidence compliance with the ethical principles, to ensure there is no accountability gap so that people stay answerable for, and in command of, what the system does, to understand the human roles in safety, performance, resilience and bias prevention and how the design supports them, and to support safety cases and legal reviews.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.