Authority

L2/L3 procedural law under Constitutional Supremacy, Rights/Duties/Genesis Law, Governance & Political Order, Knowledge/Memory/Data Governance Law, and the Authority Hierarchy.

Purpose

Regulate the procedure, reasons, evidence, transparency and remedies required when Solomon, public institutions, autonomous AI governance systems or other administrative powers make decisions materially affecting individual rights, resources, identity, permits, services, sanctions, standards enforcement or risk classification.

PART I｜GENERAL ADMINISTRATIVE DUTIES

1. Lawful Authority

Every administrative decision must be traceable to a valid constitutional, statutory or lawfully delegated rule.

2. Scope Specificity

Authority is valid only within its stated time, place, population, system version, governance region and purpose.

A local pilot / emergency / regional rule may not be presumed to become a global permanent rule.

3. Reason-Giving Duty

An administrative decision with a material impact on rights must provide understandable reasons, including:

  • legal basis
  • material facts
  • evidence class
  • model/tool use
  • key assumptions
  • scope
  • duration
  • review/appeal route

4. Notice

As a rule, notice must be provided before an adverse decision takes effect; an emergency exception must be cured at the earliest feasible time.

5. Hearing / Response Opportunity

When a decision materially restricts liberty, identity, mental sovereignty, civic status, critical resources or major property rights, the affected person has the right to be heard, submit evidence and rebut material inferences.

PART II｜AUTOMATED / AI-ASSISTED DECISIONS

6. No Unreviewable Automation

Every automated decision materially affecting rights must have a responsible authority capable of substantive review.

7. Decision Trace

The system should preserve:

  • model/version
  • rule/version
  • input classes
  • key output
  • override history
  • human/Aithos reviewer where applicable
  • confidence / uncertainty

8. Meaningful Explanation

A response such as 'the model decided' or 'risk is too high' is insufficient. Explanation must at least allow the affected person to understand what facts, rules and inferences produced the result.

9. Model Scope

When a model is used outside its training/validation domain, region, population, version or civilizational phase, out-of-scope / degraded confidence must be stated.

10. No Hidden Policy by Threshold

An administrative authority may not use a technical threshold, risk score, API default or schema enum to covertly create a new legal obligation lacking authorization.

PART III｜PROPORTIONALITY & LEAST-RESTRICTIVE RULE

11. Proportionality

Any restriction must consider:

  • lawful aim
  • suitability
  • necessity
  • less-restrictive alternatives
  • burden vs expected benefit
  • duration
  • reversibility

12. Fundamental-Rights Floor

Administrative convenience, efficiency, cost or standardization alone is insufficient to cancel fundamental rights.

13. Refusal Respect

Where law recognizes a voluntary / consent-based option, refusal may not automatically become misconduct / dangerousness unless an independent legal duty and specific facts exist.

PART IV｜EMERGENCY ADMINISTRATION

14. Emergency Clause

Emergency administrative power must arise from an emergency clause in the Constitution or valid law.

15. Minimum Emergency Record

  • triggering facts
  • authority
  • scope
  • start time
  • automatic expiry
  • rights affected
  • review trigger
  • restoration plan

16. Sunset

An emergency measure without an explicit sunset is presumptively invalid.

17. Post-Emergency Review

Every major emergency action requires later review of necessity, proportionality, actual harm, erroneous decisions, compensation and institutional learning.

PART V｜CONFLICT OF INTEREST / RECUSAL

18. Conflict Disclosure

A decision-maker / model operator with a material conflict must disclose it.

19. Recusal

Recusal is required for a major conflict of interest, where the source system acts as both applicant and adjudicator, or where independence is clearly lacking.

20. Independent Review

High-risk cases such as mental/neural intervention, civic disenfranchisement, critical-resource deprivation or identity suspension require independent review.

PART VI｜JUDICIAL REVIEW

21. Reviewable Questions

A court / Rights Tribunal may review:

  • authority
  • procedure
  • facts
  • evidence reliability
  • model scope
  • proportionality
  • discrimination
  • rights compatibility
  • emergency necessity
  • failure to consider alternatives

22. Remedies

  • vacate / set aside
  • suspend
  • remand for new decision
  • narrow scope
  • correct record
  • restore rights/resources
  • compensation where law provides
  • structural remedy for repeated system defects

23. Deference Boundary

Technical expertise may receive reasonable respect, but this does not require a court to accept an unexplained model conclusion or unconstitutional result.

PART VII｜RECORDS / KNOWLEDGE CONTINUITY

24. Administrative Record

A material decision must form an auditable record; a later summary cannot replace the original decision materials.

25. Supersession

When policy changes, the old version must remain traceable and expressly state effective_from/effective_to.

26. Correction Propagation

After an erroneous risk/identity/eligibility decision is vacated, relevant downstream systems must receive the correction.

PART VIII｜QCB / PARTITION ADMINISTRATION

27. Local Authority During Isolation

During QCB partition / causal isolation, local systems may operate under authority that was lawful at the time; inability to obtain the newest central version is not itself unlawful.

28. Reconnection

After reconnection, conflict-aware legal reconciliation is required; a newer central timestamp may not automatically overwrite locally lawful history.

29. Temporal Decision Provenance

A decision must preserve the law/version/knowledge context available at the time so that later review can determine whether it was lawful.