Selected work

Representative Papers

Moving beyond Principles: Identifying Actionable AI Fairness Practices

Apr 20, 2026

This study addresses the persistent gap between abstract ethical principles and concrete implementation in AI fairness governance, which currently lacks actionable, lifecycle-spanning guidance. Drawing on sociotechnical and practice-based perspectives, the research synthesizes 60 academic, policy, and practitioner-oriented documents and employs discourse and thematic analysis to develop the first structured, role-oriented AI fairness practice matrix. This matrix offers modular and dynamic governance guidance tailored to organizational roles and their corresponding levels of obligation, spanning the entire AI system lifecycle. By aligning responsibilities with practical actions across development, deployment, and monitoring phases, the framework provides organizations with a systematic foundation for implementing feasible and sustainable fairness governance.

0 citationsRead paper

ChatGPT's advice drives moral judgments with or without justification

Jan 03, 2025

Users increasingly rely on AI chatbots (e.g., ChatGPT) for moral decision-making, yet the psychological mechanisms underlying such reliance remain poorly understood. Method: Through controlled behavioral experiments comparing AI versus human moral advisors—with systematic variation in justification provision and source attribution—we examine adoption patterns of ethical recommendations. Contribution/Results: We provide the first empirical evidence that users adopt AI-generated moral advice not based on argument quality, but as a low-cost strategy to evade moral responsibility—termed “moral offloading.” Adoption rates show no significant difference across conditions (justified vs. unjustified advice; AI vs. human source), while AI’s accessibility amplifies this tendency. Crucially, moral offloading—not perceived technical credibility—emerges as the primary driver of advice adoption. These findings underscore the urgent need to integrate ethical literacy with digital literacy in education and policy, offering critical empirical grounding for AI ethics governance frameworks.

0 citationsRead paper

Recent publications

Latest Papers

Moving beyond Principles: Identifying Actionable AI Fairness Practices

Apr 20, 2026

This study addresses the persistent gap between abstract ethical principles and concrete implementation in AI fairness governance, which currently lacks actionable, lifecycle-spanning guidance. Drawing on sociotechnical and practice-based perspectives, the research synthesizes 60 academic, policy, and practitioner-oriented documents and employs discourse and thematic analysis to develop the first structured, role-oriented AI fairness practice matrix. This matrix offers modular and dynamic governance guidance tailored to organizational roles and their corresponding levels of obligation, spanning the entire AI system lifecycle. By aligning responsibilities with practical actions across development, deployment, and monitoring phases, the framework provides organizations with a systematic foundation for implementing feasible and sustainable fairness governance.

0 citationsRead paper

ChatGPT's advice drives moral judgments with or without justification

Jan 03, 2025

Users increasingly rely on AI chatbots (e.g., ChatGPT) for moral decision-making, yet the psychological mechanisms underlying such reliance remain poorly understood. Method: Through controlled behavioral experiments comparing AI versus human moral advisors—with systematic variation in justification provision and source attribution—we examine adoption patterns of ethical recommendations. Contribution/Results: We provide the first empirical evidence that users adopt AI-generated moral advice not based on argument quality, but as a low-cost strategy to evade moral responsibility—termed “moral offloading.” Adoption rates show no significant difference across conditions (justified vs. unjustified advice; AI vs. human source), while AI’s accessibility amplifies this tendency. Crucially, moral offloading—not perceived technical credibility—emerges as the primary driver of advice adoption. These findings underscore the urgent need to integrate ethical literacy with digital literacy in education and policy, offering critical empirical grounding for AI ethics governance frameworks.

0 citationsRead paper