“As an AI language model, I cannot”: Investigating LLM Denials of User Requests
Open accessThe DOI matches its Crossref record; 20 of the 20 references it cites were found in a registry or an index.
Abstract
Users ask large language models (LLMs) to help with their homework, for lifestyle advice, or for support in making challenging decisions. Yet LLMs are often unable to fulfil these requests, either as a result of their technical inabilities or policies restricting their responses. To investigate the effect of LLMs denying user requests, we evaluate participants’ perceptions of different denial styles. We compare specific denial styles (baseline, factual, diverting, and opinionated) across two studies, respectively focusing on LLM’s technical limitations and their social policy restrictions. Our results indicate significant differences in users’ perceptions of the denials between the denial styles. The baseline denial, which provided participants with brief denials without any motivation, was rated significantly higher on frustration and significantly lower on usefulness, appropriateness, and relevance. In contrast, we found that participants generally appreciated the diverting denial style. We provide design recommendations for LLM denials that better meet peoples’ denial expectations.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
What is AI Literacy? Competencies and Design Considerations
Long, D., & Magerko, B. (2020) · Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3313831.3376727 · Cited by 3,241 · cited 1×
“Long et al. recently defned AI literacy as an individual’s capability “to critically evaluate AI technologies" [28, p. 2], which might be challenging for novel users with a limited understanding of AI technology.”
Guidelines for Human-AI Interaction
Amershi, S., Weld, D., Vorvoreanu, M., Fourney, A., Nushi, B., Collisson, P., Suh, J., Iqbal, S., Bennett, P. N., Inkpen, K., et al. (2019) · Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3290605.3300233 · Cited by 2,439 · cited 2×
“Considering more corrective (e.g., confrontational) approaches to inappropriate user requests, such as illegal ones, is an interesting way forward for designing LLM denials and aligns with human-centred approaches that are increasingly discussed in the HCI community [1, 17, 25].”
Nass, C., Steuer, J., & Tauber, E. R. (1994) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/191666.191703 · Cited by 2,148 · cited 1×
“Following Nass’s classic work ‘Comput-ers are Social Actors’ [30], we argue that user interactions with LLMs are fundamentally social—even if users do not believe LLMs to be human or human-like.”
Douglas, B. D., Ewell, P. J., & Brauer, M. (2023) · PLOS ONE · doi:10.1371/journal.pone.0279720 · Cited by 1,532
Human heuristics for AI-generated language are flawed
Jakesch, M., Hancock, J. T., & Naaman, M. (2023) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2208839120 · Cited by 342
Communication Breakdowns Between Families and Alexa
Beneteau, E., Richards, O. K., Zhang, M., Kientz, J. A., Yip, J., & Hiniker, A. (2019) · Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3290605.3300473 · Cited by 313 · cited 1×
“This aligns with the established practice of ‘repair strategies’ in communication breakdowns [3, 24, 26].”
Co-Writing with Opinionated Language Models Affects Users’ Views
Jakesch, M., Bhat, A., Buschek, D., Zalmanson, L., & Naaman, M. (2023) · Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3544548.3581196 · Cited by 265
Tell me more?: the effects of mental model soundness on personalizing an intelligent agent
Kulesza, T., Stumpf, S., Burnett, M., & Kwan, I. (2012) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/2207676.2207678 · Cited by 225 · cited 2×
“…needs to know to successfully collaborate with an intelligent agent, indicating that those who were able to ‘improve’ their mental models of the system were more satisfied with its recommendations, suggesting that better explanations can help in aligning system output with users expectations [23].”
BackgroundResilient Chatbots: Repair Strategy Preferences for Conversational Breakdowns
Ashktorab, Z., Jain, M., Liao, Q. V., & Weisz, J. D. (2019) · Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3290605.3300484 · Cited by 224 · cited 2×
“Ashktorab et al. investigated repair strategies in human-chatbot interactions and found that participants rated a chatbot more positively when it provided alternatives and explanations (e.g., when the chatbot communicated a misunderstanding more directly and with fewer repetitions) [2].”
BackgroundAI Transparency in the Age of LLMs: A Human-Centered Research Roadmap
Liao, Q. V., & Wortman Vaughan, J. (2024) · Harvard Data Science Review · doi:10.1162/99608f92.8036d03b · Cited by 210
Automated rationale generation: a technique for explainable AI and its effects on human perceptions
Ehsan, U., Tambwekar, P., Chan, L., Harrison, B., & Riedl, M. O. (2019) · Proceedings of the 24th International Conference on Intelligent User Interfaces · doi:10.1145/3301275.3302316 · Cited by 210 · cited 1×
“The positive influence of explanations has also been highlighted by Ehsan et al., who investigated how computer agents can translate their actions into natural language explanations presented to users [14].”
BackgroundAI Transparency in the Age of LLMs: A Human-Centered Research Roadmap
Liao, Q. V., & Vaughan, J. W. (2023) · arXiv · doi:10.48550/arxiv.2306.01941 · Cited by 209 · cited 2×
“Therefore, it is important that users are provided with clear and relatable explanations [27], an aspect that was absent in baseline denials.”
BackgroundSmith-Renner, A., Fan, R., Birchfield, M., Wu, T., Boyd-Graber, J., Weld, D. S., & Findlater, L. (2020) · Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3313831.3376624 · Cited by 112 · cited 1×
“…al. found that explanations in interactive AI systems can also have negative effects: after using an intelligent text review system that provided explanations but no possibility for users to give feedback, users were significantly more frustrated than users who did not receive an explanation [34].”
BackgroundEmpathy Is All You Need: How a Conversational Agent Should Respond to Verbal Abuse
Chin, H., Molefi, L. W., & Yi, M. Y. (2020) · Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3313831.3376461 · Cited by 96
On Designing Programming Error Messages for Novices: Readability and its Constituent Factors
Denny, P., Prather, J., Becker, B. A., Mooney, C., Homer, J., Albrecht, Z. C., & Powell, G. B. (2021) · Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3411764.3445696 · Cited by 60 · cited 2×
“How we design error messages matters, as shown by Denny et al. [10], who modifed the length, language style, sentence structure, and vocabulary of error messages to improve their readability.”
Svenningsson, N., & Faraon, M. (2019) · Proceedings of the 2019 2nd Artificial Intelligence and Cloud Computing Conference · doi:10.1145/3375959.3375973 · Cited by 56 · cited 1×
“For example, Svenningsson and Faraon demonstrated that users expected a chatbot to apologize when they expressed frustration (i.e. to address their frustration and react [35]).”
BackgroundMy Bad! Repairing Intelligent Voice Assistant Errors Improves Interaction
Cuadra, A., Li, S., Lee, H., Cho, J., & Ju, W. (2021) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3449101 · Cited by 52 · cited 1×
“Their results suggest that participants perceived self-repairing VAs more positively [9].”
Do, H. J., Kong, H.-K., Tetali, P., Lee, J., & Bailey, B. P. (2023) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3579532 · Cited by 32
“I Didn’t Catch That, But I’ll Try My Best”: Anticipatory Error Handling in a Voice Controlled Game
Zargham, N., Pfau, J., Schnackenberg, T., & Malaka, R. (2022) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3491102.3502115 · Cited by 20
Machine Explanations and Human Understanding
Chen, C., Feng, S., Sharma, A., & Tan, C. (2022) · arXiv · doi:10.48550/arxiv.2202.04092 · Cited by 6
Cited by
Know Your Limits: A Survey of Abstention in Large Language Models
Wen, B., Yao, J., Feng, S., Xu, C., Tsvetkov, Y., Howe, B., & Wang, L. L. (2025) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00754 · Cited by 56
Private Yet Social: How LLM Chatbots Support and Challenge Eating Disorder Recovery
Choi, R., Kim, T., Park, S., Kim, J. G., & Lee, S.-J. (2025) · Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3713485 · Cited by 19
“A recent study examined user perceptions of such denials in various contexts, including health-related prompts [131].”
Følstad, A., Law, E. L.-C., & van As, N. (2024) · ACM Transactions on Computer-Human Interaction · doi:10.1145/3690383 · Cited by 17
LLMs Get Lost In Multi-Turn Conversation
Laban, P., Hayashi, H., Zhou, Y., & Neville, J. (2025) · arXiv · Cited by 419
“…also lacks the nuance that can occur when a human is involved in conversation, from misunderstandings over terminology, giving up due to frustration with system failures [82], or the lack of a feasible end goal for certain conversations (e.g., the user wanting a solution to an unsolved problem).”
Gnewuch, U., & Reinkemeier, F. (2026) · MIS Quarterly · doi:10.25300/misq/2025/18742 · Cited by 5
Metadata
- Authors
- Wester, J., Schrills, T., Pohl, H., & van Berkel, N.
- Venue
- Proceedings of the CHI Conference on Human Factors in Computing Systems
- Publisher
- ACM
- Published
- 2024-05-11
- Pages
- 1-14
- OpenAlex
- W4396827165
- Cited by
- 36
- References
- 20
- FWCI
- 10.95
- Type
- conference paper
- Language
- en
- Open access
- green
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.