Current approaches to agentic AI governance seem more focused on trying to apply governance after a system is developed, like a Band-Aid, instead of baking in reasonable governance and controls into the guts of the system. In the same way that security teams refer to “trust assurance” as the measures and frameworks that give them…