If the system uses continuous learning, operation and monitoring extends into continuous validation: incremental training happens while the system runs in production, so its operation is continually checked for correctness with test data, and the test data itself is updated so that it stays representative of current production data and evaluates the capabilities faithfully. Continuous validation should also feed continual improvement of the risk management processes.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.