Answers are grounded in this paper and cite their sources.
Comparative legal analysis demonstrates how explainability mandates produce exclusionary automated credit decisions, highlighting the necessity of enforceable contestability frameworks.
Automated credit scoring now structures access to consumer finance. Yet governance has centred on explainability and human oversight, often assuming that procedural transparency can secure substantive fairness. This article develops a banking-law account of how explainability and oversight can backfire when compliance is monitored through auditable artefacts rather than outcomes, yielding legible injustice: decisions that are reviewable on paper yet systematically exclusionary in effect. Using the EU model alongside US fair-lending law and the UK Consumer Duty as comparative points, it reconstructs explainability and intervention duties across borrower, governance, and supervisory audiences. The article identifies three incentive-driven mechanisms: proxy concentration, where interpretability pressures channel models toward legally familiar but socially stratified variables; defensive rationing, where accuracy-interpretability trade-offs and asymmetric liability exposure tighten credit thresholds for marginal and non-standard applicants; and review theatre, where human-in-the-loop requirements harden into workflow rituals that document responsibility without preserving real discretion. The article proposes a six-element standard for meaningful intervention: competence, authority, independence, informational sufficiency, evidential engagement and auditable records. The standard distinguishes genuine review from perfunctory click-through approval. It argues for a right of contestability, operationalised through decision-salient notice, low-friction challenge channels and credible escalation paths. It outlines outcome-facing supervision based on distributional dashboards, proxy-use diagnostics and intervention-effectiveness metrics. In that model, transparency becomes a tool for corrective enforcement rather than a device for legitimating automated lending.
No takes yet. Share an insight, caveat, or question.
Alhasan et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: