MIARO — Appendix A: Epistemic Disruptors, Non-Localizable Interventions, and the Causal Asymmetry Test

This appendix extends the MIARO framework by examining epistemic disruptions caused by non-localizable external interventions, such as opaque updates or causal intrusions inaccessible to the system’s internal history. It argues that such interventions function as asymmetry tests, revealing structural limits in origin inference even under ideal internal coherence. The analysis clarifies how rational agents may detect causal discontinuities without being able to localize or reconstruct their source, reinforcing the persistence of causal asymmetry. The appendix situates these results within debates on AI epistemology, interpretability, and the limits of self-referential explanation.