Trusting Conversational Artificial Intelligence in Mental Healthcare Provision: A Fictional Dualism Model

Paula Sweeney

doi:10.1080/15265161.2026.2645605

Trusting Conversational Artificial Intelligence in Mental Healthcare Provision: A Fictional Dualism Model

Paula Sweeney^* (Corresponding Author)

^*Corresponding author for this work

Research output: Contribution to journal › Comment/debate › peer-review

Abstract

Conversational artificial intelligence (CAI) is used in mental healthcare for prevention, intervention, psychoeducation and support. But a therapeutic relationship is surely one where trust is required: are CAIs the kinds of entities that it is appropriate to trust? Some argue that CAIs are not appropriate recipients of trust because trust requires a trustee to meet one or more condition that CAIs cannot fulfill—normative conditions such as being responsible for the outcomes of actions, acting with virtue or good character, and caring about the interests of the other. Others argue that we don’t in fact evaluate such things when we trust. Rather we trust when agents appear trustworthy and are “good players in the social game,” and CAIs can meet these conditions (Coeckelbergh Citation2012, 58).....

Original language	English
Pages (from-to)	101-103
Number of pages	2
Journal	American Journal of Bioethics
Volume	26
Issue number	5
Early online date	29 Apr 2026
DOIs	https://doi.org/10.1080/15265161.2026.2645605
Publication status	Published - May 2026

Bibliographical note

Open access via the T&F Agreement

Access to Document

10.1080/15265161.2026.2645605Licence: CC BY-NC-ND

Sweeny_AJBE_Trusting_Conversational_AI_VOR
This is an open Access article distributed under the terms of the creative commons Attribution-noncommercial-noDerivatives license (http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed,or built upon in any way. The terms on which this article has been published allow the posting of the Accepted manuscript in a repository by the author(s) or with theirconsent.
Final published version, 789 KBLicence: CC BY-NC-ND

Cite this

@article{1e6f0b805a67438b9491397e0cf41869,

title = "Trusting Conversational Artificial Intelligence in Mental Healthcare Provision: A Fictional Dualism Model",

abstract = "Conversational artificial intelligence (CAI) is used in mental healthcare for prevention, intervention, psychoeducation and support. But a therapeutic relationship is surely one where trust is required: are CAIs the kinds of entities that it is appropriate to trust? Some argue that CAIs are not appropriate recipients of trust because trust requires a trustee to meet one or more condition that CAIs cannot fulfill—normative conditions such as being responsible for the outcomes of actions, acting with virtue or good character, and caring about the interests of the other. Others argue that we don{\textquoteright}t in fact evaluate such things when we trust. Rather we trust when agents appear trustworthy and are “good players in the social game,” and CAIs can meet these conditions (Coeckelbergh Citation2012, 58).....",

author = "Paula Sweeney",

note = "Open access via the T\&F Agreement",

year = "2026",

month = may,

doi = "10.1080/15265161.2026.2645605",

language = "English",

volume = "26",

pages = "101--103",

journal = "American Journal of Bioethics",

issn = "1526-5161",

publisher = "Taylor and Francis Ltd.",

number = "5",

}

TY - JOUR

T1 - Trusting Conversational Artificial Intelligence in Mental Healthcare Provision

T2 - A Fictional Dualism Model

AU - Sweeney, Paula

N1 - Open access via the T&F Agreement

PY - 2026/5

Y1 - 2026/5

N2 - Conversational artificial intelligence (CAI) is used in mental healthcare for prevention, intervention, psychoeducation and support. But a therapeutic relationship is surely one where trust is required: are CAIs the kinds of entities that it is appropriate to trust? Some argue that CAIs are not appropriate recipients of trust because trust requires a trustee to meet one or more condition that CAIs cannot fulfill—normative conditions such as being responsible for the outcomes of actions, acting with virtue or good character, and caring about the interests of the other. Others argue that we don’t in fact evaluate such things when we trust. Rather we trust when agents appear trustworthy and are “good players in the social game,” and CAIs can meet these conditions (Coeckelbergh Citation2012, 58).....

AB - Conversational artificial intelligence (CAI) is used in mental healthcare for prevention, intervention, psychoeducation and support. But a therapeutic relationship is surely one where trust is required: are CAIs the kinds of entities that it is appropriate to trust? Some argue that CAIs are not appropriate recipients of trust because trust requires a trustee to meet one or more condition that CAIs cannot fulfill—normative conditions such as being responsible for the outcomes of actions, acting with virtue or good character, and caring about the interests of the other. Others argue that we don’t in fact evaluate such things when we trust. Rather we trust when agents appear trustworthy and are “good players in the social game,” and CAIs can meet these conditions (Coeckelbergh Citation2012, 58).....

U2 - 10.1080/15265161.2026.2645605

DO - 10.1080/15265161.2026.2645605

M3 - Comment/debate

SN - 1526-5161

VL - 26

SP - 101

EP - 103

JO - American Journal of Bioethics

JF - American Journal of Bioethics

IS - 5

ER -

Trusting Conversational Artificial Intelligence in Mental Healthcare Provision: A Fictional Dualism Model

Abstract

Bibliographical note

Access to Document

Fingerprint

Cite this