Hide or Highlight: Understanding the Impact of Factuality Expression on User Trust
Not on the map. Why?
Abstract (arXiv)
Large language models are known to produce outputs that are plausible but factually incorrect. To prevent people from making erroneous decisions by blindly trusting AI, researchers have explored various ways of communicating factuality estimates in AI-generated outputs to end-users. However, little is known about whether revealing content estimated to be factually incorrect influences users' trust when compared to hiding it altogether. We tested four different ways of disclosing an AI-generated output with factuality assessments: transparent (highlights less factual content), attention (highlights factual content), opaque (removes less factual content), ambiguity (makes less factual content vague), and compared them with a baseline response without factuality information. We conducted a human subjects research (N = 148) using the strategies in question-answering scenarios. We found that the opaque and ambiguity strategies led to higher trust while maintaining perceived answer quality, compared to the other strategies. We discuss the efficacy of hiding presumably less factual content to build end-user trust.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 23 of its 55 references that are in the map. First posted on arXiv 9 Aug 2025.
Builds on
23 of 55 references.
References · 23
- 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
- 2017A Unified Approach to Interpreting Model Predictionscited by 6,555
- 2023A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questionscited by 1,601
- 2018Don't Give Me the Details, Just the Summary! Topic-Aware Convolutional Neural Networks for Extreme Summarizationcited by 1,317
- 2023FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generationcited by 1,282
- 2020On Faithfulness and Factuality in Abstractive Summarizationcited by 900
- 2020SummEval: Re-evaluating Summarization Evaluationcited by 674
- 2020Fact or Fiction: Verifying Scientific Claimscited by 554
- 2024Language Models with Conformal Factuality Guaranteescited by 131
- 2020Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertaintycited by 129
- 2024Relying on the Unreliable: The Impact of Language Models' Reluctance to Express Uncertaintycited by 106
- 2022Generating Scientific Claims for Zero-Shot Scientific Fact Checkingcited by 59
11 more
- 2022Investigating Explainability of Generative AI for Code through Scenario-based Designcited by 47
- 2023Comparing Traditional and LLM-based Search for Consumer Choice: A Randomized Experimentcited by 36
- 2023RELIC: Investigating Large Language Model Responses using Self-Consistencycited by 23
- 2024Atomic Fact Decomposition Helps Attributed Question Answeringcited by 21
- 2021Show or Suppress? Managing Input Uncertainty in Machine Learning Model Explanationscited by 15
- 2024Designing for Appropriate Reliance: The Roles of AI Uncertainty Presentation, Initial User Decision, and User Demographics in AI-Assisted Decision-Makingcited by 11
- 2024Multi-Level Explanations for Generative Language Modelscited by 11
- 2022Trust and Reliance in XAI -- Distinguishing Between Attitudinal and Behavioral Measurescited by 10
- 2024Facilitating Human-LLM Collaboration through Factuality Scores and Source Attributionscited by 9
- 2025EvalAssist: A Human-Centered Tool for LLM-as-a-Judgecited by 5
- 2023Citations as Queries: Source Attribution Using Language Models as Rerankerscited by 1
Built on by
Citing papers per year: 2026: 1.
1 paper.