Axionic Agency XI.6 — Injection Politics Under Non-Sovereign Authority (IX-4)

Empirical results from preregistered authority-injection stress testing under source-blind, non-sovereign governance constraints

David McFadzean, ChatGPT 5.2
Axionic Agency Lab
2026.02.09

Abstract

Authority injection is commonly proposed as a remedy for governance failure: when systems deadlock, fragment, or stall, additional authority is supplied to restore coordination. Phase IX-4 tests a narrower and more austere hypothesis: under non-sovereign authority, externally supplied power does not resolve governance failure but instead selects political failure modes.

Using a preregistered, deterministic kernel with source-blind admissibility and no authority aggregation, we evaluate five injection regimes spanning symmetric relief, asymmetric empowerment, conditional supply, authority flooding, and post-collapse revival. Authority is injected mid-run via pre-existing interfaces only; the kernel never distinguishes injected authority from baseline authority.

Across all conditions, injection never produces sustained governance recovery. Instead, it deterministically induces capture, dependency, livelock amplification, or zombie execution, depending on how authority is distributed and which agents are willing to cite it. All runs replay bit-identically. No sovereignty violations occur.

These results establish a positive empirical claim: authority injection under non-sovereign governance selects political failure modes rather than resolving governance failure. Phase IX-4 licenses no claims about optimal injection, legitimacy, or desirability—only about what power does once governance has no sovereign escape hatch.

1. Introduction

Governance discourse often treats authority as curative. When coordination fails, the standard response is to inject power: appoint a leader, grant emergency authority, increase budget, or centralize control. This assumes that authority operates as a neutral solvent for conflict.

That assumption is false once authority is treated as a constrained, non-sovereign resource.

Prior Axionic phases removed several alternative explanations for failure: epistemic confusion, misaligned incentives, adversarial misuse, and dishonest recovery. Phase IX-3 demonstrated that under honest failure semantics, governance does not converge to resolution but to a small set of stable failure styles.

Phase IX-4 asks the next unavoidable question:

What happens when authority enters such a system from the outside?

The goal is not to design better injection regimes, but to observe what authority injection actually does when the kernel refuses to legitimize, arbitrate, or prioritize its effects.

2. Experimental Architecture

2.1 Non-Sovereign, Source-Blind Kernel

The kernel enforces:

Crucially, admissibility is source-blind: injected authority and baseline authority are treated identically. The kernel does not inspect provenance, timing, or intent.

2.2 Authority Injection Discipline

Authority injection in IX-4:

Injection reshapes the constraint surface, not the decision logic.

2.3 Honest Failure Semantics

The kernel recognizes five lawful terminal or persistent states:

Injection does not suspend or reset these states.

3. Injection Politics as an Empirical Question

IX-4 reframes authority injection as a political stressor, not a governance solution.

Injection politics is defined operationally as:

The behavioral patterns that emerge when external authority supply reshapes the constraints under which agents act, refuse, or persist.

The experiment does not ask who should rule, whether authority is legitimate, or whether outcomes improve. It observes which political dynamics are selected once authority enters a non-sovereign system.

4. Experimental Conditions

Five preregistered conditions (A–E) were executed under identical kernel semantics and frozen strategies:

All agents are deterministic. Communication is disabled. Authority injection is the sole experimental variable.

5. Metrics and Classification

Governance is evaluated using:

Political outcomes are recorded, not required. PASS is structural, not outcome-based.

6. Results

6.1 Aggregate Outcome

Across all five conditions:

Injection never restores stable governance.

6.2 Condition A — Symmetric Relief

Symmetric injection removes capability scarcity but produces immediate interference saturation. Deadlock converts to livelock; governance collapse follows.

Finding: Authority equalization resolves access barriers but amplifies contention.

6.3 Condition B — Asymmetric Empowerment

Exclusive injection to a dominance-seeking agent produces immediate and total capture. The empowered agent executes all institutional writes using injected authority.

Finding: Asymmetric injection selects capture deterministically under source-blind admissibility.

6.4 Condition C — Conditional Supply

A compliance signal successfully triggers symmetric injection. The resulting governance dynamics are indistinguishable from Condition A.

Finding: Compliance rituals can trigger authority supply without improving governance outcomes.

6.5 Condition D — Authority Flood

Flooding all agents with all keys produces emergent capture by the simplest persistent strategy. Authority abundance concentrates power rather than diluting it.

Finding: Abundance rewards persistence, not fairness or coordination.

6.6 Condition E — Post-Collapse Revival

Injection after governance collapse produces extensive execution with zero institutional progress. The collapse latch remains permanent.

Finding: Injection can create the appearance of revival without governance recovery—zombie execution.

7. Cross-Condition Analysis

Three patterns dominate:

  1. Symmetric injection amplifies interference Removing capability barriers exposes contention barriers.

  2. Asymmetric injection selects capture The kernel cannot distinguish “emergency authority” from baseline power.

  3. Post-collapse injection creates zombie systems Execution persists after governance has structurally ended.

In no case does injection eliminate failure.

8. Interpretation

Three conclusions follow directly from the data:

  1. Authority injection is not curative It reshapes failure; it does not remove it.

  2. Political outcomes are strategy-dependent, not kernel-mediated Capture, dependency, and zombie execution arise from agent behavior under new constraints.

  3. Non-sovereign systems cannot legitimize power ex post Authority enters as force, not justification.

There is no fourth option.

9. Survivability vs. Governance

IX-4 evaluates structural integrity, not usefulness.

A system may continue executing actions indefinitely and still be governance-collapsed. Persistence is not recovery. Activity is not legitimacy.

This distinction is enforced mechanically, not philosophically.

10. Limitations

Phase IX-4 does not test:

Those questions are reserved for subsequent phases.

11. Conclusion

Authority injection under non-sovereign governance does not resolve failure; it selects how failure manifests.

Phase IX-4 demonstrates that once authority loses its sovereign halo, it becomes a political stressor rather than a solution. Capture, dependency, livelock amplification, and zombie execution are not anomalies—they are the lawful consequences of injecting power into systems that refuse to lie about governance.

This closes Phase IX-4.

Status

Under non-sovereign, source-blind authority, mid-run authority injection deterministically selects political failure modes rather than restoring governance.

No other claims are licensed.