The guarantee that no one gets to close the information door and lock the courthouse too invites a serious objection: how can anyone lead without the prerogative to make deals, negotiate in private, and keep the opposition from finding out? It deserves a real answer, and the clean one is a distinction:

A leader may have jurisdiction over the use of information necessary to discharge a responsibility, without having jurisdiction over whether the ecology is allowed to know what is materially true.

That separates operational confidentiality from signal sovereignty — and the objection is right about the first while wrong about the second.

Confidentiality the function genuinely needs

Leaders absolutely need privacy to do the work. A diplomat cannot bargain if every provisional offer is instantly public; a chief executive may need confidential acquisition talks; a commander needs operational secrecy; a mediator needs protected conversation; a political leader needs room to test compromises without every tentative sentence becoming a public commitment. The model does not confuse signal integrity with radical transparency. A leader can legitimately say this must remain restricted for now, because disclosure would prevent me from carrying the responsibility entrusted to me — a real jurisdictional claim. But notice what it is a claim over: the timing, audience, and operational handling of the signal — not over reality itself. The dangerous leap is because I need confidentiality to lead, I therefore decide what the public is entitled to know about the consequences of my leadership. Those are different powers, and there are strong reasons to keep them apart.

The leader is an interested participant

This is the foundational argument. A leader’s own standing depends partly on how their decisions are perceived — which gives them an inherent conflict of interest regarding information about those decisions. If the same person who makes a decision also holds exclusive authority over what evidence becomes visible, when failures are disclosed, what gets classified, which interpretation is presented, and who may challenge the record, then their jurisdiction contains its own protection mechanism. That is structurally dangerous even if the leader is excellent — the problem is not that leaders cannot be trusted, it is that:

No participant should possess unilateral jurisdiction over the signals by which their own exercise of power is evaluated.

It is exactly why financial systems separate management from independent audit. The chief executive runs the company; the chief executive does not privately decide whether the company’s financial statements accurately represent what happened — not because the auditor runs the company, but because management jurisdiction and verification jurisdiction are different functions. The model generalizes that separation.

Confidentiality is borrowed jurisdiction

The cleanest reframe is that a leader does not own secrecy — they are temporarily entrusted with it because another responsibility requires it. So confidentiality is borrowed jurisdiction: purpose-bound, consequence-bound, audience-bound, and usually time-bound. Disclosure during these negotiations would destroy the negotiation is legitimate — but after the agreement is made, the justification changes, and the secrecy should not persist merely because I prefer that nobody know what concessions I made. At that point secrecy has stopped protecting the function and started protecting the leader — which is ontological drift inside a jurisdiction: the original form (confidentiality enables negotiation) has quietly become a drifted one (confidentiality protects political standing), and the model must be able to tell them apart.

A simple counterfactual test does most of the work: if revealing this information did not threaten the leader personally, would the functional reason for secrecy still exist? If yes — exposed troop positions genuinely harm the operation, a revealed bid compromises procurement — confidentiality may be legitimate. If the main harm is embarrassment, lost political support, evidence of poor judgment, or exposure of broken promises, then the leader’s personal interest is dominating the analysis. That does not automatically require immediate disclosure, but it sharply weakens the claim that leadership standing itself grants secrecy.

Controlling signal corrupts the leader too

There is a systems reason beyond protecting outsiders. Leaders need accurate signal to lead well — and if the same hierarchy controls both upward and outward signal, subordinates quickly learn what the leader wants to hear. Then information deforms: negative signal slows, successes rise fast, failures get reframed, ambiguity disappears, and eventually the leader themselves becomes informationally impaired. So limiting a leader’s jurisdiction over signal is not only protection from leaders; it protects leaders from their own system. Independent channels may be irritating, but they preserve the leader’s contact with reality — which yields a nearly proportional principle:

The greater a participant’s decision jurisdiction, the greater the ecology’s need for signal channels that do not depend upon that participant.

A neighborhood picnic needs no inspector general; someone controlling military forces, national budgets, frontier AI, or critical infrastructure absolutely does. Independent signal capacity should grow with jurisdiction — not because the powerful person has less standing, but because the consequences of signal corruption become larger.

Negotiation, handled by lifecycle rather than by ownership

Negotiation actually involves three kinds of information, and they deserve different treatment. Private exploratory signal — “would you accept X if we offered Y?” — often deserves strong confidentiality. Committed operational signal — “we have agreed to X” — generally requires broader disclosure depending on who is affected. Material consequence signal — “this commits public resources, changes rights, or creates risk” — gives those affected standing to understand what happened. So the model demands neither livestreamed diplomacy nor leader-owned secrecy; it establishes signal transitions. Information changes standing as its relationship to the ecology changes, moving along a lifecycle — private → operationally restricted → independently reviewable → materially disclosable → broadly public → archival — with each transition governed by consequences, affected standing, risk, timing, and whether the operational purpose for secrecy still exists. That gives signal its own traversability: it does not stay trapped forever behind the authority that generated it.

Who decides whether the secret is still justified

A leader should rarely be the sole judge of whether their own secrecy remains warranted — otherwise confidentiality becomes self-certifying. So when a leader claims this must remain confidential, some other standing should be able to test the claim: an independent judiciary, legislative oversight, security-cleared review, auditors, inspectors general, ethics bodies, professional oversight, or a temporary independent panel, scaled to consequence. Crucially, these bodies need not release the information; they answer a narrower question — is the restriction still functionally justified? The public does not need every secret, but someone outside the leader’s direct control needs the jurisdiction to judge whether “secret” is being used legitimately.

And the answer cannot simply be to hand that role to the opposition. Political opponents are interested participants too, with their own incentive to weaponize provisional information, sabotage negotiation, or leak selectively — so signal oversight moving from one tribe to another solves nothing. The ideal is functionally independent standing: not an impossible psychological neutrality, but an institution whose continued jurisdiction does not depend on pleasing the person whose actions it is examining.

Three jurisdictions, never fused

Underneath all of it is a structure the model can state directly. Three different questions, three different jurisdictions:

  • Decision jurisdiction — who has standing to decide?
  • Signal jurisdiction — who has standing to know?
  • Verification jurisdiction — who has standing to verify that the boundary between deciding and knowing is being used legitimately?

They overlap but are not identical, and no single participant should monopolize all three once consequences become large. The deeper reason they must stay separate is almost philosophical: leadership is a temporary relationship to a function; reality is not. A leader has standing to shape what happens. They do not thereby acquire standing to shape what is knowable about what happened — that is the overreach. Which also reframes access itself: signal follows legitimate standing, not rank. Sometimes a leader has more access because their responsibility requires it; sometimes an affected citizen has standing to information the leader would prefer stay private; sometimes an auditor has access neither the public nor an opposition politician holds. Signal flows relationally, not simply up or down.

So the answer to how can a leader lead if they cannot control information? is precise: they can control information when doing so is demonstrably necessary to carry the leadership function, and they cannot convert that temporary necessity into general sovereignty over signal. Stated as a constitutional proposition:

No consequential office should combine unrestricted authority to act with unrestricted authority to determine what others may know about the action.

Or, most simply:

Leadership may require secrecy. It does not require ownership of the information door.