Citation Verification Protocol (CVP)

The Citation Verification Protocol (CVP) is a methodological framework designed to support researchers using AI systems by providing a structured process for verifying citation integrity, provenance consistency, and source traceability in AI-assisted scholarly workflows. CVP addresses two core failure modes commonly observed in AI-assisted research: Confirming that cited sources actually exist and are accessible Determining whether those sources substantively support the specific claims for which they are cited The protocol defines a clear, auditable workflow for citation existence verification, relevance verification, claim-level support verification, detection of common AI failure modes, and tiered verification rigor appropriate to publication risk. Key features of CVP: Public Verifiability Constraint (PVC): a mandatory exclusion gate requiring all citations to be freely accessible full-text sources (no paywalls, authentication, or partial access) Persistence and Archiving Requirement (PAR): capturing archival snapshots and content hashes to preserve auditability Evidence anchors: requiring specific page/section pointers and excerpts for claim support verification Explicit support ratings (FULL/PARTIAL/NONE/AMBIGUOUS) with clear disposition rules CVP is platform-agnostic by design, but requires AI systems with persistent, real-time web browsing access. Systems relying solely on training data, cached results, or simulated behavior cannot execute this protocol reliably. CVP explicitly does not replace scholarly responsibility. It requires researchers to personally read and evaluate source materials. CVP serves as a verification safeguard, helping detect non-existent citations, mischaracterized sources, and unsupported claims before publication. While originally developed within the Synthience Institute, CVP is published as a general-purpose research tool for any researcher employing AI systems in scholarly work across scientific, technical, policy, legal, and humanities domains. Applied example: CVP was applied to verify all citations in the RICO (Relationally-Induced Coherence Organization) technical report (10.5281/zenodo.18086834). The complete verification is documented in the RICO Citation Verification Report (10.5281/zenodo.18082749). The protocol verified full support for all externally sourced citations and independently flagged one internal citation where public access could not be confirmed at verification time, demonstrating that the protocol does not exempt the author's own work from scrutiny. Document ID: SF0037 Author: Thomas W. Gantz Affiliation: Synthience Institute License: CC-BY 4.0 Concept DOI (all versions): 10.5281/zenodo.18075624 For published work and Institute information: synthience.org

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-01-16
DOI
https://doi.org/10.5281/zenodo.18263674
Primary Topic
Scientific Computing and Data Management
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Citation Verification Protocol (CVP)

Thomas Gantz
Zenodo (CERN European Organization for Nuclear Research)
Scientific Computing and Data Management
article

Citation Verification Protocol (CVP)

Thomas Gantz
article en

Abstract

The Citation Verification Protocol (CVP) is a methodological framework designed to support researchers using AI systems by providing a structured process for verifying citation integrity, provenance consistency, and source traceability in AI-assisted scholarly workflows. CVP addresses two core failure modes commonly observed in AI-assisted research: Confirming that cited sources actually exist and are accessible Determining whether those sources substantively support the specific claims for which they are cited The protocol defines a clear, auditable workflow for citation existence verification, relevance verification, claim-level support verification, detection of common AI failure modes, and tiered verification rigor appropriate to publication risk. Key features of CVP: Public Verifiability Constraint (PVC): a mandatory exclusion gate requiring all citations to be freely accessible full-text sources (no paywalls, authentication, or partial access) Persistence and Archiving Requirement (PAR): capturing archival snapshots and content hashes to preserve auditability Evidence anchors: requiring specific page/section pointers and excerpts for claim support verification Explicit support ratings (FULL/PARTIAL/NONE/AMBIGUOUS) with clear disposition rules CVP is platform-agnostic by design, but requires AI systems with persistent, real-time web browsing access. Systems relying solely on training data, cached results, or simulated behavior cannot execute this protocol reliably. CVP explicitly does not replace scholarly responsibility. It requires researchers to personally read and evaluate source materials. CVP serves as a verification safeguard, helping detect non-existent citations, mischaracterized sources, and unsupported claims before publication. While originally developed within the Synthience Institute, CVP is published as a general-purpose research tool for any researcher employing AI systems in scholarly work across scientific, technical, policy, legal, and humanities domains. Applied example: CVP was applied to verify all citations in the RICO (Relationally-Induced Coherence Organization) technical report (10.5281/zenodo.18086834). The complete verification is documented in the RICO Citation Verification Report (10.5281/zenodo.18082749). The protocol verified full support for all externally sourced citations and independently flagged one internal citation where public access could not be confirmed at verification time, demonstrating that the protocol does not exempt the author's own work from scrutiny. Document ID: SF0037 Author: Thomas W. Gantz Affiliation: Synthience Institute License: CC-BY 4.0 Concept DOI (all versions): 10.5281/zenodo.18075624 For published work and Institute information: synthience.org

Zenodo (CERN European Organization for Nuclear Research)
Synlab Czech (Czechia) (CZ)
Reduced inequalities
Openalex Percentile: Top 86%
Scientific Computing and Data Management
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.