The Illusion of Independent Quorums: Epistemic Fault Domains and Correlated Cognitive Failures in Agentic Quorums

Summary

Formalizes why agent quorums that look independent are not: distinct reviewers share upstream telemetry, documents and tool backends, so many votes can collapse onto a single corrupted root cause - replication does not imply epistemic redundancy. The authors define Epistemic Fault Domains and a Structural Epistemic Cut that lower-bounds the number of root faults needed for semantic compromise, and prove that arbitrarily large quorums can retain a cut of 1 and that adding voters at a fixed threshold cannot increase credited resilience. This attacks the conditional-independence assumption on which Sah and Stiglitz's polyarchy and committee results rest, giving the ISR paper a formal criterion for when a voting aggregation rule delivers only fake redundancy.

Contribution

Queued — not yet read.

Method

Queued — not yet read.

Relevance to RISE

Surfaced by the 2026-09-07 literature sweep as a high-priority candidate for the Architecture as Epistemology (ISR) paper (angle: aggregating LLM judgments; correlated errors / homogeneity across LLM agents; organization theory applied to AI agents; epistemic quality of LLM outputs; MAS failure and cost analyses). Verified against: Verified via arXiv API metadata record on 2026-09-07: full abstract, 2 authors, v1, published 2026-08-24, primary cs.DC (also cs.MA, cs.SE), 13 pages, code and frozen benchmark at github.com/openkedge/efd.

Critique / open questions

Queued — not yet assessed.

Key quotes

Multi-agent quorums are widely used to authorize high-stakes infrastructure and policy mutations, yet distinct reviewers often share upstream telemetry, documents, or tool backends. (abstract)

When upstream inputs fail, multiple votes collapse onto a single corrupted cause: replication does not imply epistemic redundancy. (abstract)