Advanced control principles for external connections (required, with baseline and intermediate, of an AI whose inherent risk is high). The security implications of every change to external or third-party connections are validated and documented beforehand; the institution works with providers to maintain and improve the security of those connections, for example end-to-end encryption and leased lines.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.