The influence of mental state attributions on trust in large language models

Abstract Rapid advances in artificial intelligence (AI) have led users to believe that systems such as large language models (LLMs) have mental states, including the capacity for ‘experience’ (e.g., emotions and consciousness). These folk-psychological attributions often diverge from expert opinion...

Full description

Saved in:
Bibliographic Details
Main Authors: Clara Colombatto, Jonathan Birch, Stephen M. Fleming
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:Communications Psychology
Online Access:https://doi.org/10.1038/s44271-025-00262-1
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850124444942467072
author Clara Colombatto
Jonathan Birch
Stephen M. Fleming
author_facet Clara Colombatto
Jonathan Birch
Stephen M. Fleming
author_sort Clara Colombatto
collection DOAJ
description Abstract Rapid advances in artificial intelligence (AI) have led users to believe that systems such as large language models (LLMs) have mental states, including the capacity for ‘experience’ (e.g., emotions and consciousness). These folk-psychological attributions often diverge from expert opinion and are distinct from attributions of ‘intelligence’ (e.g., reasoning, planning), and yet may affect trust in AI systems. While past work provides some support for a link between anthropomorphism and trust, the impact of attributions of consciousness and other aspects of mentality on user trust remains unclear. We explored this in a preregistered experiment (N = 410) in which participants rated the capacity of an LLM to exhibit consciousness and a variety of other mental states. They then completed a decision-making task where they could revise their choices based on the advice of an LLM. Bayesian analyses revealed strong evidence against a positive correlation between attributions of consciousness and advice-taking; indeed, a dimension of mental states related to experience showed a negative relationship with advice-taking, while attributions of intelligence were strongly correlated with advice acceptance. These findings highlight how users’ attitudes and behaviours are shaped by sophisticated intuitions about the capacities of LLMs—with different aspects of mental state attribution predicting people’s trust in these systems.
format Article
id doaj-art-a7f33ca40b714447ad15131c0da7d7e8
institution OA Journals
issn 2731-9121
language English
publishDate 2025-05-01
publisher Nature Portfolio
record_format Article
series Communications Psychology
spelling doaj-art-a7f33ca40b714447ad15131c0da7d7e82025-08-20T02:34:19ZengNature PortfolioCommunications Psychology2731-91212025-05-01311710.1038/s44271-025-00262-1The influence of mental state attributions on trust in large language modelsClara Colombatto0Jonathan Birch1Stephen M. Fleming2Department of Psychology, University of WaterlooDepartment of Philosophy, Logic and Scientific Method, and Centre for Philosophy of Natural and Social Science, London School of Economics and Political ScienceDepartment of Experimental Psychology and Max Planck UCL Centre for Computational Psychiatry and Ageing Research, University College LondonAbstract Rapid advances in artificial intelligence (AI) have led users to believe that systems such as large language models (LLMs) have mental states, including the capacity for ‘experience’ (e.g., emotions and consciousness). These folk-psychological attributions often diverge from expert opinion and are distinct from attributions of ‘intelligence’ (e.g., reasoning, planning), and yet may affect trust in AI systems. While past work provides some support for a link between anthropomorphism and trust, the impact of attributions of consciousness and other aspects of mentality on user trust remains unclear. We explored this in a preregistered experiment (N = 410) in which participants rated the capacity of an LLM to exhibit consciousness and a variety of other mental states. They then completed a decision-making task where they could revise their choices based on the advice of an LLM. Bayesian analyses revealed strong evidence against a positive correlation between attributions of consciousness and advice-taking; indeed, a dimension of mental states related to experience showed a negative relationship with advice-taking, while attributions of intelligence were strongly correlated with advice acceptance. These findings highlight how users’ attitudes and behaviours are shaped by sophisticated intuitions about the capacities of LLMs—with different aspects of mental state attribution predicting people’s trust in these systems.https://doi.org/10.1038/s44271-025-00262-1
spellingShingle Clara Colombatto
Jonathan Birch
Stephen M. Fleming
The influence of mental state attributions on trust in large language models
Communications Psychology
title The influence of mental state attributions on trust in large language models
title_full The influence of mental state attributions on trust in large language models
title_fullStr The influence of mental state attributions on trust in large language models
title_full_unstemmed The influence of mental state attributions on trust in large language models
title_short The influence of mental state attributions on trust in large language models
title_sort influence of mental state attributions on trust in large language models
url https://doi.org/10.1038/s44271-025-00262-1
work_keys_str_mv AT claracolombatto theinfluenceofmentalstateattributionsontrustinlargelanguagemodels
AT jonathanbirch theinfluenceofmentalstateattributionsontrustinlargelanguagemodels
AT stephenmfleming theinfluenceofmentalstateattributionsontrustinlargelanguagemodels
AT claracolombatto influenceofmentalstateattributionsontrustinlargelanguagemodels
AT jonathanbirch influenceofmentalstateattributionsontrustinlargelanguagemodels
AT stephenmfleming influenceofmentalstateattributionsontrustinlargelanguagemodels