1Kosmos Proposes Delegated Authority Model for AI Governance

Michael Engle, cofounder of 1Kosmos, argues that the traditional human-in-the-loop model of AI governance is becoming a bottleneck for organizations. He asserts that requiring humans to approve individual actions one by one is inefficient given the volume and velocity of decisions made by autonomous agents in enterprise workflows.

Engle proposes a transition toward governing delegated authority. This model focuses on the continuous validation that an AI agent operates within the specific authority granted by a verified human at runtime.

This approach moves beyond static authentication. It allows organizations to explain why a specific action was permitted based on active policies and delegated permissions, rather than focusing primarily on why access was denied.