01 · Authority and attribution
Competence
Legal authority for the act and for the way the system participates.
Public instrument
Intelligent Institutional Act Test
Eight gates. One accountable public act. No aggregate score. A small applied expression of the programme, not the whole of it; not an operational system; not a compliance product.
- For
- Global institutional leadership: those who authorise, review, audit, adjudicate or correct intelligence-mediated public power.
- Unit
- One dated institutional act · named body · identified legal position.
- Status
- Public instrument; findings, not scores.
- Standing
- Small applied expression of the programme. Not the whole of JUS.automica, not an operational system, not a compliance product.
Who exercised public power, under what authority, through what record, and with what route to challenge and correction?
After an act has occurred, accuracy and procurement questions do not settle legitimacy. The Test converts the question above into an eight-gate examination of one intelligence-mediated institutional act, designed so that readers in Australian, Dutch, British, digital-public-infrastructure, ombuds and tribunal settings can recognise their own institutional acts: not a Pakistan-only frame, and not “global compliance”.
What it is not
The limits are part of the instrument.
Not AI-compliance software · not a country ranking or legality score · not certification · not a substitute for judicial determination · not governmental authority or operational deployment · not IGF endorsement · not the whole identity of JUS.automica.
NOT-ESTABLISHED means missing evidence, not an accusation of failure. If the act, the body, the legal position and the system’s role cannot be named, the Test is not yet runnable. Scheme-level scandals may teach structure; they are not, by themselves, completed act-level findings.
The eight gates
Dualities must not be collapsed.
Gate questions follow the Global Observatory Proposal Guide, design edition 2.0 (2026). No gate establishes lawfulness by itself.
02 · Authority and attribution
Protected Institutional Judgment
Entrusted evaluative judgment remains where the law requires it.
03 · Authority and attribution
Adoption
Lawful adoption of the operative rule, model, threshold or workflow and its applicable version; informed adoption of the individual output and reasons.
04 · Authority and attribution
Attribution
Traceable attribution to a responsible institution and office. “The system decided” is not an answer.
05 · Review, evidence and remedy
Reconstructability
The material sequence from inputs and rules to reasons, action and notice can be reproduced.
06 · Review, evidence and remedy
Contestability
Timely notice, meaningful reasons, access to material information, a usable opportunity to respond.
07 · Review, evidence and remedy
Independent Counter-Capacity
A body independent of the original path can obtain the record and grant effective relief.
08 · Review, evidence and remedy
Correction
Repair of the individual outcome and of the upstream rule, data or workflow that generated recurrent error.
Findings
Four findings, recorded separately from the evidence.
One jurisdictional defect cannot compensate for another. A missing record must not be disguised as a weak pass.
- PASS
- Reliable evidence establishes the gate is satisfied under the applicable legal standard.
- FAIL
- Reliable evidence establishes the gate is not satisfied.
- NOT-ESTABLISHED
- The record supports neither pass nor fail; the missing evidence remains explicit.
- NOT-APPLICABLE
- The gate does not apply to this act, with reasons recorded.
Continuity of legality
What the Test pressure-tests, at act level.
The five thesis doctrines of the Doctrine of Intelligent Legal Order.
- Unbroken Authority Chain
- Computational Due Process
- Ratio-Preservation
- Legal-Computational Fidelity
- Sovereign Control Requirement
Comparative horizon · teaching only
Scheme-level teaching, explicitly not completed one-act findings: Robodebt (AU) · SyRI (NL) · Toeslagen risk triage (NL) · MiDAS (US-MI). Dated-act-preferring units: Sisbén IV classification acts (CO) · Horizon prosecutions, Hamilton-style (UK) · Huduma Namba / NIIMS (KE) as identity-infrastructure horizon. No PASS/FAIL tables as verdicts, no rankings. These are not verified Observatory records.
Run the Test
The Intelligence Engine conducts the Test with you: it asks for the unit of analysis, examines each gate in turn and records the eight findings to your account, with no score.
Examine an act