Intelligent systems do not break when they fail. They break when they continue to function and no one can justify why authority is being exercised. This is not a technical failure. It is a failure of legitimacy. Before any action, there is a prior question: may this proceed legitimately at all? Modern systems answer after execution. This declaration restores what must come before. The Prudential Era begins when intelligence remains operational after legitimacy has already failed. At sufficient scale and delegation, optimization ceases to produce resilience and instead amplifies systemic fragility. From that point onward, legitimacy cannot be repaired ex post. It must be preserved ex ante. This declaration establishes a normative threshold, not a technical solution. Authority must remain explicitly authorized, legible, and interruptible before action is permitted to occur. It does not seek to control intelligence. It preserves the conditions under which governance, responsibility, and refusal remain possible. Where those conditions cannot be sustained, authorization must be withdrawn.