The DOI matches its Crossref record; 9 of the 9 references it cites were found in a registry or an index.
People choose to receive human empathy despite rating AI empathy higher
Abstract
Recent advances in AI have enabled large language models to produce expressions that seem empathetic to human users, raising scientific and ethical questions about how people perceive and choose between human and AI sources of emotional support. Although an increasing number of studies have examined how people rate empathy generated by AI, little to no work has examined whether people would choose to receive empathy from AI. We conducted four studies investigating whether people prefer to receive empathetic expressions from humans or AI, and how they evaluate these expressions. Across diverse samples and stimuli, we found evidence for what we term the AI empathy choice paradox: participants significantly preferred to receive empathy from humans, yet they rated AI-generated empathetic responses as higher in quality, more effective at making them feel heard, and more effortful when they did choose them. These findings contribute to ongoing debates about AI empathy by demonstrating that while people may avoid AI as an empathy source, they nonetheless benefit from AI empathy when they experience it. Our results suggest potential applications for AI in supplementing human emotional support while highlighting the importance of respecting individual preferences for empathy sources.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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
Algorithm aversion: People erroneously avoid algorithms after seeing them err.
Dietvorst, B. J., Simmons, J. P., & Massey, C. (2015) · Journal of Experimental Psychology: General · doi:10.1037/xge0000033 · Cited by 3,232 · cited 2×
“Wealsonotethatour fi nding of AI empathy avoidance may not be speci fi c to empathy per se, but rather a symptom of a larger algorithmic aversion 34,35 .”
Ayers, J. W., Poliak, A., Dredze, M., Leas, E. C., Zhu, Z., Kelley, J. B., Faix, D. J., Goodman, A. M., Longhurst, C. A., Hogarth, M., et al. (2023) · JAMA Internal Medicine · doi:10.1001/jamainternmed.2023.1838 · Cited by 2,613 · cited 5×
“While AI applications in coding, healthcare, and data analysis have gained acceptance 1,2 the emergence of AI in emotional and relational domains has sparked intense debate among researchers and ethicists alike 3,4 .”
InfluentialCausal diagrams for empirical research
PEARL, J. (1995) · Biometrika · doi:10.1093/biomet/82.4.669 · Cited by 2,355 · cited 1×
“These DAGs identify the set of control variables that must be included in each model for appropriateinference of causal effects(i.e., thesuf fi cientadjustment set) 30,31 .”
People are averse to machines making moral decisions.
Bigman, Y. E., & Gray, K. (2018) · Cognition · doi:10.1016/j.cognition.2018.08.003 · Cited by 738
Large Language Model Influence on Diagnostic Reasoning: A Randomized Clinical Trial
Goh, E., Gallo, R., Hom, J., Strong, E., Weng, Y., Kerman, H., Cool, J. A., Kanjee, Z., Parsons, A. S., Ahuja, N., et al. (2024) · JAMA Network Open · doi:10.1001/jamanetworkopen.2024.40969 · Cited by 724
Durably reducing conspiracy beliefs through dialogues with AI
Costello, T. H., Pennycook, G., & Rand, D. G. (2024) · Science · doi:10.1126/science.adq1814 · Cited by 354
AI can help people feel heard, but an AI label diminishes this impact
Yin, Y., Jia, N., & Wakslak, C. J. (2024) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2319112121 · Cited by 247
AI will never convey the essence of human empathy
Perry, A. (2023) · Nature Human Behaviour · doi:10.1038/s41562-023-01675-w · Cited by 162
Artificial empathy in healthcare chatbots: Does it feel authentic?
Seitz, L. (2024) · Computers in Human Behavior: Artificial Humans · doi:10.1016/j.chbah.2024.100067 · Cited by 137
Research works printed but not linked: 28
- Mitchell, M. The metaphors of artificial intelligence. Science 386, eadt6140 (2024). · DOI
- Zaki, J. & Ochsner, K. N. The neuroscience of empathy: progress, pitfalls and promise. Nat. Neurosci. 15, 675–680 (2012). · DOI · Paper page
- Shteynberg, G. et al. Does it matter if empathic AI has no empathy? Nat. Mach. Intell. 6, 496–497 (2024). · DOI · Paper page
- Inzlicht, M., Cameron, C. D., D’Cruz, J. & Bloom, P. In praise of empathic AI. Trends Cogn. Sci. 28, 89–91 (2024). · DOI
- Wenger, J. D., Cameron, C. D., Orlandi, M. & Inzlicht, M. What’s so special about human empathy? Examining AI empathy and its tradeoffs. PsyArXiv https://doi.org/10.31234/osf.io/69kzb_v1 (2025). To appear in empathy and artificial intelligence: challenges, advances, and ethical considerations (eds. Cameron, C. D. & Perry, A.). · doi:10.31234/osf.io/69kzb_v1
- Ovsyannikova, D., de Mello, V. O. & Inzlicht, M. Third-party evaluators perceive AI as more compassionate than expert humans. Commun. Psychol. 3, 4 (2025). · DOI
- Rubin, M. et al. Comparing the value of perceived human versus AI-generated empathy. Nat. Hum. Behav. 9, 2345–2359 (2025). · DOI
- Faul, F., Erdfelder, E., Buchner, A. & Lang, A. G. Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav. Res. Methods 41, 1149–1160 (2009). · DOI · Paper page
- Cameron, C. D., Scheffer, J. A., Hadjiandreou, E. & Anderson, S. Motivated empathic choices. In Advances in Experimental Social Psychology 66, 191–279 (Elsevier, 2022). · DOI
- Keysers, C. & Gazzola, V. Dissociating the ability and propensity for empathy. Trends Cogn. Sci. 18, 163–166 (2014). · DOI · Paper page
- Zaki, J. Empathy: a motivated account. Psychol. Bull. 140, 1608–1647 (2014). · DOI · Paper page
- Cameron, C. D. et al. Empathy is hard work: people choose to avoid empathy because of its cognitive costs. J. Exp. Psychol. Gen. 148, 962–976 (2019). · DOI · Paper page
- Bruneau, E., Dufour, N. & Saxe, R. How we know it hurts: item analysis of written narratives reveals distinct neural responses to others’ physical pain and emotional suffering. PLoS ONE 8, e63085 (2013). · DOI
- Scheffer, J. A., Cameron, C. D. & Inzlicht, M. Caring is costly: people avoid the cognitive work of compassion. J. Exp. Psychol. Gen. 151, 172–196 (2022). · DOI
- Depow, G. J., Francis, Z. & Inzlicht, M. The experience of empathy in everyday life. Psychol. Sci. 32, 1198–1213 (2021). · DOI · Paper page
- Wickham, H. ggplot2: Elegant Graphics for Data Analysis (Version 3.5.1) [Software]. https://ggplot2.tidyverse.org (GGPLOT, 2016). · DOI · Paper page
- Kuznetsova, A., Brockhoff, P. B. & Christensen, R. H. B. lmertest package: tests in linear mixed effects models. J. Stat. Softw. 82, 1–26 (2017). · DOI · Paper page
- Elwert, F. Graphical Causal Models in Handbook of Causal Analysis for Social Research (ed. Morgan, S. L.), 245–273 (Springer, 2013). · DOI · Paper page
- Wysocki, A. C., Lawson, K. M. & Rhemtulla, M. Statistical control requires causal justification. Adv. Methods Pract. Psychol. Sci. 5, 1–19 (2022).
- Flora, D. B. Your coefficient alpha is probably wrong, but which coefficient omega is right? A tutorial on using R to obtain better reliability estimate. Adv. Methods Pract. Psychol. Sci. 3, 484–501 (2020). · DOI · Paper page
- Shiverdecker, L. K. & LeBreton, J. M. A primer on multilevel (random coefficient) regression modeling in The Handbook of Multilevel Theory, Measurement, and Analysis (eds. Humphrey, S. E. & LeBreton, J. M.), 389–422 (American Psychological Association, 2019). · doi:10.1037/0000115-018
- Danaher, J. The rise of the robots and the crisis of moral patiency. AI Soc. 34, 129–136 (2019). · DOI
- Reinecke, M. G., Kappes, A., Porsdam Mann, S., Savulescu, J. & Earp, B. D. The need for an empirical research program regarding human–AI relational norms. AI Ethics 5, 71–80 (2025). · DOI
- Böhm, R., Jörling, M., Reiter, L. & Fuchs, C. People devalue generative AI’s competence but not its advice in addressing societal and personal challenges. Commun. Psychol. 1, 32 (2023). · DOI
- Osborne, M. R. & Bailey, E. R. Me vs. the machine? Subjective evaluations of human- and AI-generated advice. Sci. Rep. 15, 3980 (2025). · DOI
- Myers, S. & Everett, J. A. C. People expect artificial moral advisors to be more utilitarian and distrust utilitarian moral advisors. Cognition 256, 106028 (2025). · DOI
- de Mello, V. O.et al. The Moralization of Artificial Intelligence. PsyArXiv 10.31234/osf.io/5mwre_v5 (2025). · doi:10.31234/osf.io/5mwre_v2
- Shamay-Tsoory, S. G. & Hertz, U. Adaptive empathy: a model for learning empathic responses in response to feedback. Perspect. Psychol. Sci. 17, 1008–1023 (2022). · DOI
Metadata
- Authors
- Wenger, J. D., Cameron, C. D., & Inzlicht, M.
- Venue
- Communications Psychology
- Published
- 2026-01-31
- Volume
- 4
- Issue
- 1
- ISSN
- 2731-9121
- Cited by
- 18
- References
- 9
- Language
- en
- Licence
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor, 4 d)
- Funders
- National Science Foundation (DGE1255832)
- Gouvernement du Canada | Natural Sciences and Engineering Research Council of Canada (RGPIN-2019-05280)
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: People choose to receive human empathy despite rating AI empathy higher. https://app.cochat.ai/papers/10.1038/s44271-025-00387-3
- The paper
- Wenger, J. D., Cameron, C. D., & Inzlicht, M. (2026). People choose to receive human empathy despite rating AI empathy higher. Communications Psychology. https://doi.org/10.1038/s44271-025-00387-3
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.
