When implementing: set correctly the points to note in using the AI system or service; use AI within the range set by the developer; secure at the time of provision the accuracy of the system and, where needed, the recency of training data; and examine whether the business user's environment differs from the use environment the developer envisaged. After provision starts: periodically verify that the system or service is used for proper purposes.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.