Before adopting AI the body starts from the problem to solve and the end-user's needs, and with a cross-functional team including data specialists and domain experts weighs: the alternative solutions and their advantages, disadvantages and cost against budget, including simpler methods that could give the same quality faster or cheaper; whether the available data is accurate, representative, complete and large enough; whether the data can be used responsibly and lawfully under the GDPR; whether benefits outweigh prospective negative outcomes; whether all users benefit equally or some gain at others' cost; whether the required skills are available; and whether it will solve the problem, with the metrics that will test that.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.