Nothing Decided in the Dark

Isaac Aluochier, S.Arb, S.Adj, Arbitral Fellow, CPM · Aluochier Dispute Resolution

Published July 2026 · a commentary responding to a live public debate at time of writing

In recent weeks, arguments that used to stay inside law firms and registries have spilled into public view. How does an urgent application move from filing on a Friday night to a signed order by the following morning? Who is told, and when? Does the written record of what happened actually match what happened? These are not new questions. What is new is that Kenyans are now asking them out loud, in public, about their own courts — and finding that the honest answer, in ordinary litigation, is often: nobody outside the room can be entirely sure.

That uncertainty is not a moral failing unique to any one judge or any one case. It is a structural feature of how ordinary judicial procedure was built. Interim relief has always depended on urgency, and urgency has always created a gap between when a decision is made and when the affected party learns of it. The gap is not new. What deserves attention is that a different, stronger safeguard against exactly this gap already exists in Kenyan law — just not in the courtroom.

Section 4(3) of the Fair Administrative Action Act imposes an obligation that has no real equivalent in ordinary court procedure. Before any administrative decision-maker takes action likely to adversely affect a person's rights, that person must be given adequate prior notice of the nature and purpose of the proposed action, and a reasonable opportunity to make representations — before the decision is made, not after. Not eventual notice through a court filing. Not the chance to appeal once the order already exists. Notice of the proposal itself, in advance, with a genuine opportunity to change the decision-maker's mind.

Ordinary litigation was never built this way, because it relies on a different logic — the adversarial contest of pleadings and hearings is presumed, on its own, to deliver fairness. That presumption holds well for matters heard in the ordinary course. It holds far less well for urgent, one-sided applications decided overnight, exactly the pattern now drawing public scrutiny. If Section 4(3)'s obligation applied to that pattern the way it applies to administrative action, an order could not lawfully be signed without the affected party first being told, in writing, precisely what was proposed against them, and given a real chance to be heard on it before it was made final.

This is not a hypothetical improvement ADR aspires to. It is how these tribunals are constitutionally required to operate, because their proceedings are administrative action within the meaning of Article 47 and the Fair Administrative Action Act. Before any Determination or Award that would adversely affect a party is issued, that party must first receive adequate written notice of exactly what is proposed — the finding, the evidence behind it, the legal reasoning, and the relief to be ordered — and must be given a genuine opportunity to respond before the decision is finalised. In AIETAR electoral proceedings this takes the form of a published notice with a fixed response window measured in hours, not weeks, because electoral timelines demand it — but the underlying obligation is the same one that applies across every AITAR and AISTAR proceeding.

An order made in the small hours, unseen by the party it affects until it is already in force, is not a shortcut this framework permits under pressure of urgency. It is not available at all. The notice obligation cannot be dispensed with by urgency, by agreement between the parties, or by analogy to a different procedure that happens to be silent on it. It is a precondition to a lawful decision, and its omission is, on its own, a ground for the decision to be set aside on review — regardless of whether the underlying outcome was otherwise correct.

A second, quieter form of the same problem is the question of how a matter reaches a particular decision-maker in the first place — whether by transparent, rules-based process, or by informal contact between one party and the person who will decide their case. In these frameworks, that question has a structural answer rather than a discretionary one. Matters are allocated through the ADR Registry against published criteria — track, tier eligibility, conflict checks — not through a party or their counsel reaching the adjudicator directly. There is no version of “who alerted the tribunal member to the filing” to be asked here, because the allocation mechanism is not a private conversation. It is a documented Registry function, open to the same scrutiny as everything else these institutions do.

A framework can promise fair process and still fail if the person applying it answers to nobody. That is why Supervisory Review Tribunal oversight in these frameworks applies without exception — including to the Chief Adjudicator's own Determinations, reviewed by the SRT on precisely the same terms as any other Tribunal member's. There is no tier of decision in AITAR, AISTAR, or AIETAR practice that is, by virtue of who made it, placed beyond that review. The question “who reviews the reviewer” has a published, standing answer here, not a discretionary one decided case by case.

Every Determination and Award issued under these frameworks is entered, permanently, in the AITAR Digital Gazette — not sealed in a registry file accessible only on request, not subject to a clerk's discretion about who may see it. A decision that will be read is a decision that must be capable of being defended on its own terms, in public, indefinitely. That permanence is not a courtesy extended to the curious. It is the last and most durable of the safeguards this framework is built on — because a process that is transparent by design has nothing that depends on being taken on trust.

None of what is described here is a pledge about how ADR intends to behave. It is a description of mechanisms already written into the AITAR, AISTAR, and AIETAR instruments — the pre-decision notice obligation, rules-based Registry allocation, universal SRT review, and permanent Gazette publication — each independently checkable against the published rules, none of them dependent on any individual's good character to hold. That is a deliberately higher bar than a promise, because promises can be broken quietly and mechanisms cannot be dispensed with without the omission itself becoming a ground for review.

Kenya's disputants do not have to choose between hoping a decision-maker behaves well and having no forum at all. A framework built so that the safeguards do not depend on anyone's good behaviour already exists, and it is receiving matters now.