What Makes an AI System Human-Centered? Preliminary Findings from an Empirical Study

This study presents an empirical investigation into perceptions of human-centered artificial intelligence (HCAI) by analyzing qualitative responses from 136 AI practitioners, academics, and students. While AI systems are increasingly integrated into everyday life, a critical gap persists between theoretical HCAI frameworks and empirically derived, stakeholder-informed guidelines. To address this gap, participants were asked: “What makes an AI system human-centered?” Thematic analysis using affinity diagramming revealed seven core themes: Ethics and Privacy, User-Centric Design, Transparency and Explainability, Human Augmentation, Emotional Intelligence, Inclusivity, and Societal Responsibility. Among these, Ethics and User-Centric Design were most frequently emphasized, suggesting strong alignment between practitioner insights and existing theoretical models. This study contributes to HCAI research by identifying empirically grounded principles that support the development of AI systems that are ethical, transparent, inclusive, and responsive to human needs. The findings offer actionable guidance for AI developers, policymakers, and educators committed to advancing human-centered approaches that align technological innovation with societal values.