CoChat
Sign Up
Cited by 15Artificial Intelligence in Healthcare and EducationHealth, Technology, Consumer BehaviorTelemedicine and Telehealth Implementation

Influencing Public Acceptance of Artificial Intelligence (AI) in Healthcare Delivery

Open access
Aras, Selin, Drakos, Calvin, Manimangalam, Vineesha, Nasir, Moiz Ali, Burns, Christina, Smith, Davey, Equils, Ozlem·2026·Frontiers in Digital Health
DOI 10.3389/fdgth.2025.1664345PMC PMC12853031
Open this paper in CoChat — free accountOpen PDF

Abstract

Introduction: Despite the potential of artificial intelligence (AI) to transform healthcare delivery and reduce costs, adoption remains uneven across populations. Understanding the demographic, behavioral, and cognitive factors influencing public willingness to use AI-powered health tools is critical for equitable implementation. This study examined determinants of AI adoption in healthcare among adults in the United States (U.S.). Methods: A cross-sectional survey was conducted between March and June 2024 using convenience sampling across the U.S. The study included 568 adult respondents recruited via Qualtrics. The survey assessed demographic characteristics, digital health behaviors, self-reported health status, cognitive and attitudinal factors, and behavioral intentions related to AI use in healthcare. Logistic regression models were used to examine associations between predictors and willingness to adopt AI, with z-tests for subgroup comparisons and Bonferroni correction applied for multiple hypothesis testing. Results: The sample was predominantly female (66.7%) and Hispanic/Latino (50.7%), with moderate income and education levels. Older age was negatively associated with AI adoption ( β = −0.029), males were less likely to use AI than females ( β = −0.388), and income was positively correlated with AI adoption ( β = 0.096). Trust in AI was substantially lower than trust in physicians: 14.6% trusted ChatGPT's diagnosis for serious illness compared with 92.3% trusting physicians, and 17.1% versus 96.4% for specialist referrals. Telehealth use strongly predicted AI adoption ( β = 1.012), while lower self-rated mental health was associated with higher AI use ( β = −0.254). Uninsured participants reported higher trust in AI diagnostic capabilities than insured participants (57% vs. 43%, p < 0.05). Ethnic differences were observed, with Asian participants reporting higher AI usage rates than Hispanic participants (16.49% vs. 5.56%, p < 0.05). Discussion: AI adoption in healthcare is shaped by the interaction of demographic, socioeconomic, and cultural factors. While AI has the potential to expand healthcare access, adoption patterns reflect existing disparities in healthcare access and trust. Trust emerged as a central determinant, with AI functioning as a compensatory tool when traditional healthcare access is limited. Given the U.S.-specific context, findings should be interpreted as exploratory and may not generalize to other healthcare systems. These results highlight the need for future research on transparency, digital literacy, and structural barriers to support equitable implementation of healthcare AI.

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 · 10.3389/fdgth.2025.1664345

Cites

32 of 35 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?

  • Willingness to use AI tools varies by demographic and socioeconomic status, with lower income populations accepting these technologies less often.

    “Willingness to use AI tools varies by demographic and socioeconomic status, with lower income populations accepting these technologies less often, which complicates equitable implementation [10].”

    From Artificial Intelligence and Machine Learning for Identifying Social Determinants of Health in Low‐Income Populations Within United States Health Systems: A Scoping Review (2026) · cites this paper as [10]

    The Results section ('Demographic factors and AI adoption') reports differences in AI use by age, gender, education, and race or ethnicity. It does not establish that lower-income participants accepted AI less often: the income association was not statistically significant (p = 0.39). The paper measures use and adoption, which are not identical to willingness to use AI. (LLM, conf=0.88)

    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

Research works printed but not linked: 4

  1. [16]Siau K Wang W. Building trust in artificial intelligence, machine learning, and robotics. Cutter Bus Technol J. (2018) 31(2): 47– 53. Available online at: https://www.cutter.com/article/building-trust-artificial-intelligence-machine-learning-and-robotics-498981 (Accessed September 22, 2025).
  2. [17]Leventhal H Brissette I Leventhal EA. The common-sense model of self-regulation of health and illness. In: Cameron LD Leventhal H, editors. The Self-Regulation of Health and Illness Behaviour. London: Routledge (2003). p. 42– 65.
  3. [23]Agresti A. Statistical Methods for the Social Sciences. 5th ed. Boston: Pearson (2018).
  4. [30]Hosmer DW Lemeshow S Sturdivant RX. Applied Logistic Regression. 3rd ed. Hoboken, NJ: Wiley (2013).

Cited by

Sorted by citations

Related

Authors
Aras, S., Drakos, C., Manimangalam, V., Nasir, M. A., Burns, C., Smith, D., & Equils, O.
Affiliations
  • MiOra
  • University of California System (US)
Corresponding author
Equils, Ozlem
ORCID
Venue
Frontiers in Digital Health
Publisher
Frontiers Media SA
Published
2026-01-13
Volume
7
Pages
1664345-1664345
ISSN
2673-253X (ISSN-L)
DOI
10.3389/fdgth.2025.1664345
PMC
PMC12853031
OpenAlex
W7123357850
Cited by
15
References
32
Citations by year
2026: 15
FWCI
29.50
Type
journal article
Language
en
Update notices
correction (2026-07-02)
Open access
Licence
Locations
  • Frontiers in Digital Health (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • PubMed Central (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
  • Health Sciences › Health Professions › Health Information Management › Health, Technology, Consumer Behavior
  • Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Telemedicine and Telehealth Implementation
Keywords
artificial intelligence adoption, artificial intelligence in healthcare, public acceptance, trust in AI, trust in physicians, telehealth utilization, self-rated mental health, insurance status, socioeconomic disparities, ethnic differences, digital health behavior, behavioral intention, equitable implementation, digital literacy, health care access disparities, United States adults, gender differences, age differences

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