This study examines how artificial intelligence (AI) digital employee integration (AIDEI) influences organizational adaptability (OA) via cross-boundary knowledge sharing (CBKS) and collective psychological safety (CPS), and how polarization about AI-related job displacement (PAIJD) moderates these links.
We conducted a multi-method study of 405 banking managers in China, Europe and the United States, plus 36 interviews. The analyses included partial least squares structural equation modeling, fuzzy-set qualitative comparative analysis and importance-performance map analysis to capture complex socio-technical dynamics.
AIDEI positively predicts CBKS and – contrary to surveillance arguments – CPS; both mediate AIDEI's effect on OA. Polarization about AI-related job displacement amplifies the positive relationships between AI digital employee integration and both CBKS and CPS. CPS shows accelerating marginal returns to OA (significant quadratic term).
The single-industry, time-lagged design limits causal and cross-sector inference; longitudinal, multi-sector studies are warranted. The findings extend social identity theory (SIT) to AI-intensive work by specifying when integration enhances, rather than erodes, relational resources.
Organizations should prioritize cultivating CPS when implementing AI digital employees, as it offers the highest leverage for enhancing OA. Managers should channel workplace polarization about AI constructively rather than suppressing it, recognizing its potential to energize adaptive responses.
This research integrates SIT with IT affordances to explain why AIDEI can broaden superordinate identity and depersonalize error attribution, thereby enabling CBKS and CPS; we identify threshold effects for CPS and equifinal configurations of OA.
