Human–AI Collaboration in Data-Driven Decision-Making: Trust Calibration, Explainability, and Decision Quality

Recent advances in artificial intelligence (AI), machine learning, and generative models have accelerated the adoption of AI-assisted decision-making across domains such as finance, healthcare, cybersecurity, and public administration. While these technologies offer substantial improvements in predictive accuracy and efficiency, growing concerns have emerged regarding algorithmic opacity, automation bias, fairness, accountability, and human overreliance on AI recommendations.As a result, research has increasingly shifted from fully autonomous systems toward human–AI collaborative approaches, where human judgment and machine intelligence jointly contribute to decision outcomes. Despite rapid progress, important questions remain regarding how trust should be calibrated, when AI assistance improves decision quality, and under which conditions collaboration may negatively influence performance. Addressing these challenges requires interdisciplinary research integrating technical, behavioral, organizational, and ethical perspectives.This Research Topic aims to investigate the conditions under which human–AI collaboration improves or impairs decision quality in data-driven environments. Although hybrid intelligence systems are often promoted as superior to either human-only or AI-only decision-making, empirical evidence remains fragmented regarding when these benefits are realized and what factors influence decision effectiveness.The central objective is to examine how trust calibration, explainability, human oversight, and behavioral responses shape the outcomes of collaborative decision processes. Particular attention will be paid to risks such as automation bias, excessive reliance on AI systems, diminished human vigilance, and algorithm aversion, whereby users reject valid AI recommendations. To address these challenges, the Research Topic encourages rigorous empirical and methodological studies comparing human-only, AI-only, and hybrid decision-making approaches. By identifying the mechanisms that support effective collaboration, this collection seeks to develop evidence-based principles for designing trustworthy, transparent, and accountable AI systems that enhance decision quality while maintaining meaningful human control, particularly in high-stakes contexts.This Research Topic welcomes theoretical, empirical, methodological, and interdisciplinary contributions examining the effectiveness of human–AI collaboration in decision-making. We particularly encourage studies that compare human-only, AI-only, and human–AI collaborative approaches using measurable outcomes such as decision accuracy, trust calibration, cognitive workload, fairness, accountability, error detection, and user acceptance.Topics of interest include, but are not limited to:Trust calibration and reliance on AI recommendationsAutomation bias and algorithm aversionExplainable AI (XAI) and interpretable decision-support systemsHuman-in-the-loop and human-on-the-loop frameworksHybrid intelligence and augmented decision-makingHuman–AI collaboration in high-stakes domains, including finance, healthcare, cybersecurity, and public policyFairness, transparency, governance, and accountability in AI-supported decisionsExperimental and behavioral studies evaluating human–AI performanceWe welcome Original Research Articles, Review Articles, Methods Papers, Perspectives, and interdisciplinary studies that advance understanding of effective human–AI collaboration and its impact on decision quality. Topic editors: Asefeh Asemi, Dr. Sayyed Khawar Abbas, Ali Alibeigi Source: https://www.frontiersin.org/research-topics/82778/human-ai-collaboration-in-data-driven-decision-making-trust-calibration-explainability-and-decision-quality