Epistemic Infrastructure Assurance: A Statutory Framework for Metric Integrity in Public Decision-Making

This policy note identifies a specific governance gap in the use of research metrics within public and academic decision-making. Citation counts, download statistics, h-index scores, rankings, and related indicators are increasingly used as inputs into consequential decisions concerning promotion, remuneration, funding allocation, and institutional standing. Yet the integrity, provenance, auditability, and contestability of these metrics remain weakly assured. The note develops the concept of epistemic infrastructure to describe systems whose outputs function as decision-relevant signals, and introduces a regulatory standard of reliability-for-reliance: outputs are fit for use where they are sufficiently accurate, auditable, and contestable to justify institutional reliance without independent re-verification. It argues that integrity failures in such systems are often silent—arising from internal defects, structural design features, or unauthorised manipulation—while systems remain operational and outputs appear plausible. Drawing on the governance lessons of the Post Office Horizon scandal, the note highlights the risks of treating system-generated outputs as de facto truth in the absence of effective verification and challenge mechanisms. It analyses the existing legal baseline in UK law, including UK GDPR (in particular Articles 5, 28, and 32), administrative law duties of inquiry, and contractual obligations of good faith, and concludes that the issue lies not in the absence of legal standards but in the absence of operational assurance practices. The note proposes a set of Epistemic Infrastructure Assurance Duties—including public posture disclosures, independent audit, auditability and logging, provenance and data lineage, contestability mechanisms, and incident notification—implemented primarily through UK GDPR Article 28 processor obligations and procurement conditions by UKRI and related bodies. It identifies a significant remedial gap for individuals affected by inaccurate or unreliable metrics and argues that targeted primary legislation establishing contestability rights and an adjudicatory mechanism is required to close that gap. The proposed framework provides an immediate, legally grounded pathway for improving the integrity of evaluative metric systems, while setting out the case for longer-term statutory reform.