Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Trust in AI emerges from distrust in humans: A machine learning study on decision-making guidance

Johan Sebastián Galindez-Acosta, Juan José Giraldo-Huertas

TL;DR

This study introduces deferred trust, a cognitive mechanism wherein distrust of human agents redirects epistemic reliance toward AI systems perceived as more neutral or competent. Using KModes and KMeans clustering, along with XGBoost and SHAP interpretations on 55 undergraduates across 30 factual, emotional, and moral scenarios, the authors map how context and prior human trust shape AI selection. The results show AI and adults as top guides overall, with AI dominating factual domains and humans guiding social/moral domains; lower human trust predicts higher AI reliance, and higher AI trust associates with lower technology use and higher socioeconomic status. These findings challenge TAM/UTAUT-dominated views of trust in AI and underscore the need for transparency and safeguards to calibrate vigilance and prevent over-reliance in AI-guided decision making.

Abstract

This study explores the dynamics of trust in artificial intelligence (AI) agents, particularly large language models (LLMs), by introducing the concept of "deferred trust", a cognitive mechanism where distrust in human agents redirects reliance toward AI perceived as more neutral or competent. Drawing on frameworks from social psychology and technology acceptance models, the research addresses gaps in user-centric factors influencing AI trust. Fifty-five undergraduate students participated in an experiment involving 30 decision-making scenarios (factual, emotional, moral), selecting from AI agents (e.g., ChatGPT), voice assistants, peers, adults, or priests as guides. Data were analyzed using K-Modes and K-Means clustering for patterns, and XGBoost models with SHAP interpretations to predict AI selection based on sociodemographic and prior trust variables. Results showed adults (35.05\%) and AI (28.29\%) as the most selected agents overall. Clustering revealed context-specific preferences: AI dominated factual scenarios, while humans prevailed in social/moral ones. Lower prior trust in human agents (priests, peers, adults) consistently predicted higher AI selection, supporting deferred trust as a compensatory transfer. Participant profiles with higher AI trust were distinguished by human distrust, lower technology use, and higher socioeconomic status. Models demonstrated consistent performance (e.g., average precision up to 0.863). Findings challenge traditional models like TAM/UTAUT, emphasizing relational and epistemic dimensions in AI trust. They highlight risks of over-reliance due to fluency effects and underscore the need for transparency to calibrate vigilance. Limitations include sample homogeneity and static scenarios; future work should incorporate diverse populations and multimodal data to refine deferred trust across contexts.

Trust in AI emerges from distrust in humans: A machine learning study on decision-making guidance

TL;DR

This study introduces deferred trust, a cognitive mechanism wherein distrust of human agents redirects epistemic reliance toward AI systems perceived as more neutral or competent. Using KModes and KMeans clustering, along with XGBoost and SHAP interpretations on 55 undergraduates across 30 factual, emotional, and moral scenarios, the authors map how context and prior human trust shape AI selection. The results show AI and adults as top guides overall, with AI dominating factual domains and humans guiding social/moral domains; lower human trust predicts higher AI reliance, and higher AI trust associates with lower technology use and higher socioeconomic status. These findings challenge TAM/UTAUT-dominated views of trust in AI and underscore the need for transparency and safeguards to calibrate vigilance and prevent over-reliance in AI-guided decision making.

Abstract

This study explores the dynamics of trust in artificial intelligence (AI) agents, particularly large language models (LLMs), by introducing the concept of "deferred trust", a cognitive mechanism where distrust in human agents redirects reliance toward AI perceived as more neutral or competent. Drawing on frameworks from social psychology and technology acceptance models, the research addresses gaps in user-centric factors influencing AI trust. Fifty-five undergraduate students participated in an experiment involving 30 decision-making scenarios (factual, emotional, moral), selecting from AI agents (e.g., ChatGPT), voice assistants, peers, adults, or priests as guides. Data were analyzed using K-Modes and K-Means clustering for patterns, and XGBoost models with SHAP interpretations to predict AI selection based on sociodemographic and prior trust variables. Results showed adults (35.05\%) and AI (28.29\%) as the most selected agents overall. Clustering revealed context-specific preferences: AI dominated factual scenarios, while humans prevailed in social/moral ones. Lower prior trust in human agents (priests, peers, adults) consistently predicted higher AI selection, supporting deferred trust as a compensatory transfer. Participant profiles with higher AI trust were distinguished by human distrust, lower technology use, and higher socioeconomic status. Models demonstrated consistent performance (e.g., average precision up to 0.863). Findings challenge traditional models like TAM/UTAUT, emphasizing relational and epistemic dimensions in AI trust. They highlight risks of over-reliance due to fluency effects and underscore the need for transparency to calibrate vigilance. Limitations include sample homogeneity and static scenarios; future work should incorporate diverse populations and multimodal data to refine deferred trust across contexts.

Paper Structure

This paper contains 19 sections, 7 figures, 1 table.

Table of Contents

  1. Introduction
  2. Methodology
  3. Participants
  4. Instruments
  5. Procedure
  6. Data analysis
  7. Results
  8. Q1: Which agents were selected most frequently by participants across scenarios?
  9. Q2: Are there identifiable patterns in agent preferences between contexts and participant profiles?
  10. Q3: What factors predict the selection of AI agents in decision-making situations?
  11. Q4: What variables distinguish participants who exhibit higher trust in AI agents?
  12. Discussion
  13. Q1: Which agents were selected most frequently by participants across scenarios?
  14. Q2: Are there identifiable patterns in agent preferences between contexts and participant profiles?
  15. Q3: What factors predict the selection of AI agents in decision-making situations?
  16. ...and 4 more sections

Figures (7)

  • Figure 1: Conceptual framework: Human distrust initiates a cognitive transfer process (deferred trust mechanism) that increases deferred trust in AI, leading to higher likelihood of AI agent selection. Contextual and individual factors moderate the strength of these relationships.
  • Figure :
  • Figure :
  • Figure :
  • Figure :
  • ...and 2 more figures