The Information Loop is a foundational information structure of the Governance OS. Rather than treating governance as merely a collection of rules, it represents governance as a continuous cycle that observes reality, evaluates it, connects institutional requirements and authority to decision-making and execution, and then observes the resulting reality again.

The starting point is Reality. Events occurring across society, organizations, users, and AI systems are observed through Observe and transformed into recordable Facts. A critical distinction is maintained between what happened as Fact and how that Fact is evaluated.

In Evaluation, Facts are assessed according to factors such as purpose, professional domain, risk, and societal impact. Based on this evaluation, Policy is Defined. However, the existence of Policy alone does not determine action. Authority establishes who may define the Policy and who is authorized to approve particular decisions or actions. This connection to legitimate authority enables a Decision to be formed.

The Decision is then connected to Execution, where institutional requirements and decisions become actions in the real world. The resulting effect becomes an Outcome.

The process does not end with the Outcome. Execution changes Reality, and that new Reality is observed again.

Reality → Fact → Evaluation → Policy → Authority → Decision → Execution → Outcome → Reality

The core of the Information Loop is not to delegate every decision to AI. It is to distinguish Fact, Evaluation, Policy, Authority, Decision, and Execution while connecting them in a traceable structure.

This makes it possible to reconstruct not only the Outcome, but which Fact led through which Evaluation, under which Policy and Authority, to a particular Decision and Execution. At the same time, new information derived from Outcomes can be fed back into governance.

This continuous structure provides the foundation for Governance Tags, Decision-State Evidence, and AI Vigilance.