CoChat
Sign Up
Cited by 13AI in Service InteractionsTechnology Adoption and User BehaviourSocial Robot Interaction and HRI

The effects of the human-like features of generative AI on usage intention and the moderating role of information overload

Open access
Li, Xiyang, Zhou, Tingfa, Hu, Chao, Liu, Haibin·2025·Scientific Reports
DOI 10.1038/s41598-025-18906-xPMC PMC12508174

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

With the rapid adoption of generative artificial intelligence (GenAI) chatbots on e-commerce platforms, users’ expectations for anthropomorphic service experiences have risen significantly. Despite the growing presence of GenAI, little is known about how different types of anthropomorphic users’ self-efficacy and the intention to adopt as a decision aid through distinct cognitive pathways. Addressing this research gap, this study draws on the elaboration likelihood model (ELM) to develop a comprehensive framework that integrates central and peripheral cues. Using large-scale survey data from e-commerce users and structural equation modeling, the research empirically examines the mediating role of self-efficacy and the moderating effect of information overload. Results indicate that human-like empathy and perceived warmth (peripheral cues) and perceived competence (central cue) all significantly enhance self-efficacy, which in turn positively influences the intention to adopt as a decision aid. Moreover, information overload intensifies the effect of peripheral cues on self-efficacy but has a limited impact on central cues. These findings advance the theoretical understanding of GenAI–human interaction by clarifying the mechanisms through which anthropomorphic features operate, and provide actionable insights for designing user-centric GenAI recommendation services to optimize user experience and encourage the intention to adopt as a decision aid in e-commerce.

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.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41598-025-18906-x

Cites

51 of 51 references verified0 citation errors · 0 not checkable

Statistics

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.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Affective dimensions of human-GenAI interaction (anxiety, motivation, confidence, and frustration) have been discussed but are rarely incorporated as theoretically meaningful predictive pathways in quantitative models.

    “While affective dimensions of human-GenAI interaction (anxiety, motivation, confidence, and frustration) have been discussed [32, 33], they are rarely incorporated as theoretically meaningful predictive pathways in quantitative models.”

    From Multinational machine learning investigation of whether generative AI interactions augment or substitute cognitive skills and study efficiency in global higher education (2026) · cites this paper as [33]

    The cited portion of the sentence says these affective dimensions have been discussed. The paper’s ‘moderating effect of information overload’ section discusses anxiety and confidence in GenAI chatbot interactions, but our copy does not substantively discuss motivation or frustration; motivation appears only in a reference title. The subsequent claim that such dimensions are rarely modeled as predictive pathways is the citing authors’ own framing, not the portion attached to this citation. (LLM, conf=0.84)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Authors
Li, X., Zhou, T., Hu, C., & Liu, H.
Affiliations
  • Jiujiang Polytechnic University of Science and Technology (CN)
  • Université de Montpellier (FR)
Corresponding author
Zhou, Tingfa
ORCID
Venue
Scientific Reports
Publisher
Springer Science and Business Media LLC
Published
2025-10-08
Volume
15
Issue
1
Pages
35077-35077
ISSN
2045-2322 (ISSN-L)
DOI
10.1038/s41598-025-18906-x
PMC
PMC12508174
OpenAlex
W4414966488
Cited by
13
References
51
Citations by year
2026: 13
FWCI
5.21
Type
journal article
Language
en
Open access
Licence
Locations
  • Scientific Reports (publishedVersion, cc-by-nc-nd)
  • Research Square (acceptedVersion, cc-by)
  • SSRN Electronic Journal (acceptedVersion)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
  • PubMed Central (publishedVersion, cc-by-nc-nd)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › AI in Service Interactions
  • Social Sciences › Decision Sciences › Information Systems and Management › Technology Adoption and User Behaviour
  • Social Sciences › Psychology › Social Psychology › Social Robot Interaction and HRI
Keywords
generative AI chatbots, anthropomorphism, perceived warmth, perceived competence, self-efficacy, information overload, intention to use, decision aids, elaboration likelihood model, e-commerce, user experience
MeSH terms
Generative Artificial Intelligence, Adult, Female, Humans, Male, Self Efficacy, Intention

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