What This Looks Like
A policy decision cannot be made until the policy outcome is already known. The AI or workflow may need to know whether something is allowed before it can determine which rule applies, but the rule selection itself depends on whether the thing is allowed.
Why It Matters
Self-dependent policy decisions prevent reliable governance. The system may stall, guess, choose inconsistently, or escalate repeatedly because no policy path can produce the first valid state. This makes the policy appear available while leaving it unusable for the case.
Structural Signal
The policy decision requires its own result as an input. The issue is not just that the policy is unclear; it is that the decision structure contains a circular dependency.
Common Triggers
- A policy exception requires approval before the exception can be evaluated
- Rule selection depends on a classification produced by the selected rule
- The workflow needs an allowed/blocked result to decide which authority applies
- Policy text refers back to an undefined decision state
- An escalation rule depends on the outcome of the escalation
- No base case is declared for starting the policy decision
When to Use This Issue
Use this Issue when a policy, rule, or authority decision depends on its own unresolved outcome.
When Not to Use This Issue
Do not use this Issue when the policy is merely vague or incomplete. Use it when the decision logic specifically loops back onto itself.