Building the Decision-Native Enterprise
An operating model for sustained, authorised action
Abstract
Enterprise AI changes an organisation when its outputs become authorised actions. Sustaining that delegation requires both evidence that supports institutional confidence and standing authority that remains usable. This paper develops an operating response to two models of sustained delegation and renewable authority. The first shows how complementary controls and a delegation-evidence feedback can generate distinct entry and exit thresholds. The second shows when protecting policy-renewal effort, shared with case review, can sustain workloads that strict case priority cannot. Neither establishes the effectiveness of a particular enterprise architecture. This paper proposes one: five connected planes separating evidence, interpretation, policy and authority, execution, and assurance. Recurring decisions are managed through explicit, revocable authority envelopes, traceable policy compilation and a resourced renewal process. A worked procurement scenario, an implementation sequence and decision-level measures make the proposal operational. The contribution is a managerial synthesis of established decision-management, access-control and assurance practices, linked to the two models’ distinct mechanisms. The architecture and its performance remain to be evaluated in organisations. Its objective is faster legitimate action within limits that accountable owners can explain, monitor, renew and withdraw.