Processor Fidelity
Sixth trust facet — how faithfully the judge reads evidence at the bind
Processor fidelity is the sixth facet of trustworthiness — the judge-side complement to the five artifact-side facets (tier, commitment, authority weight, atomicity, epistemic uncertainty). It measures how faithfully the human or collective at the commitment boundary reads tier, uncertainty, and evidence at commit time, without systematic distortion.
Can I rely on this judge, now? — parallel to artifact uncertainty’s can I rely on this instance, now?
Processor fidelity does not replace tier or artifact coordinates. It modulates how much of the artifact’s real trustworthiness survives contact with the decision room. Low processor fidelity is a discount factor on the composed judgment — the mirror image of high epistemic uncertainty on the artifact. See composition rules in The Facets Composed.
Typical collapses: fatigued three-second approval of an L1-U-med input; rubber-stamp approval despite perfect calibration weights; tier inflation driven by disposition rather than evidence.
Raise fidelity with bind mechanisms from Strengthening the Commitment Boundary — blind review, stress testing, override tracking — when outcome severity warrants. Full index of processor failure modes: Disposition and Processor Fidelity.