This study examines how agentic leadership, conscious unbossing and AI-augmented ethical decision-making (AI-EDM) jointly shape crisis-resilient leadership among emerging global leaders. Addressing a key gap in leadership scholarship, the study conceptualizes artificial intelligence (AI) not merely as a technical tool but as a governance architecture that structures decision authority, accountability and ethical action under conditions of uncertainty.
A mixed-method design was employed. Survey data were collected from 423 early-career professionals across 12 countries. Partial least squares structural equation modeling using SmartPLS (version 4.1.1.6) tested a moderated mediation model in which agentic leadership predicts crisis-resilient leadership effectiveness (CR-LE) through AI-EDM, with conscious unbossing and crisis exposure as moderating conditions. Qualitative interviews with 24 Gen Z leaders operating in crisis-prone contexts were used to contextualize and interpret the quantitative findings.
Agentic leadership significantly predicted CR-LE, with this relationship mediated by AI-EDM. The indirect pathway was stronger among leaders exhibiting higher levels of conscious unbossing, indicating that leadership agency is conditioned by decentralized and algorithmically governed authority structures. Crisis exposure intensified the salience of ethical AI engagement, highlighting the heightened accountability pressures leaders face under volatility. Qualitative findings reinforced these results, illustrating how leaders navigate responsibility for outcomes shaped by AI-mediated decision systems they do not fully control.
The cross-sectional design limits causal inference, and scalar measurement invariance was not tested. Future longitudinal and cross-contextual research is needed to examine how leadership agency and accountability operate across different institutional and regulatory environments.
The findings underscore the need to reconceptualize leadership development and organizational governance for AI-mediated contexts. Rather than focusing solely on ethical capability enhancement, organizations and policymakers must equip leaders to operate under algorithmic constraints by clarifying decision rights, escalation pathways and contestability mechanisms.
This study advances leadership theory by identifying accountability displacement under AI-mediated governance as a central mechanism reshaping leadership agency in crisis contexts. It reframes leadership not as autonomous discretion or technological mastery but as accountable action under constraint within socio-technical governance systems.
