Verified Misguidance: Measuring Structural Citation Failures in Search-Augmented LLMs

2026-05-27

AI Summary

This study addresses a systemic flaw in the citation structures of current retrieval-augmented large language models, which often produce “confirmatory misinformation”—citations that are factually accurate yet structurally distorted, thereby misleading users without their awareness. To tackle this issue, the authors introduce the CITETRACE dataset and propose the first generalizable three-dimensional evaluation framework that comprehensively assesses citation quality across intent alignment, source appropriateness, and answer faithfulness. Leveraging large-scale real-world queries, expert rating matrices, a five-point faithfulness scale, and cross-model citation chain tracing, the analysis reveals that 30.6% of citations distort original sources, 27.1% stem from domain-mismatched references, and 96% of users encounter at least one instance of structural misrepresentation. Citation quality is found to be predominantly influenced by differences among model providers.

Abstract

Users of search-augmented LLMs rely on citations as evidence that responses are grounded in real sources, and rarely verify the cited pages themselves. Millions of queries per day now pass through these systems, making citation quality a silent determinant of whether users are informed or misled-yet existing benchmarks each address one facet in isolation, leaving the joint structure that determines citation trustworthiness unmeasured. We construct CITETRACE, a large-scale dataset that traces the full citation chain from user query through retrieved source to generated answer: 11,200 real-world queries from 28 communities paired with 112,000 responses from ten models across five providers, yielding 761,495 evaluable citation pairs. We design a three-dimension evaluation framework that scores each citation on intent-purpose alignment, source suitability, and answer-source fidelity, using expert-validated predefined matrices and a five-level fidelity rubric; the framework applies to any system that produces citation-bearing responses. Applying this framework at scale, we identify a systematic pattern we call VERIFIED MISGUIDANCE (VM): models cite real, accessible sources yet fail along one or more dimensions, producing a fidelity-suitability trade-off in which faithful models select inappropriate sources and vice versa. Across our pool, 30.6% of citations distort their sources and 27.1% originate from domain-inappropriate sources; at the response level, up to 96% of users encounter at least one structurally misleading citation. Provider-level differences explain 88-96% of citation-quality variance, suggesting that source selection is governed more by factors beyond individual model capability than by the LLMs themselves. Together, CITETRACE and its evaluation framework provide the first resource for diagnosing structural citation failures in deployed search-augmented systems.

Problem

• citation failure
• search-augmented LLMs
• verified misguidance
• structural trustworthiness
• citation quality

Innovation

• Verified Misguidance
• CITETRACE
• citation fidelity
• search-augmented LLMs
• structural citation failure