Information Governance Reconfigures Crisis Information Ecologies: An Agent-Based Model of Trust Collapse, Rumor Dominance, and Polarization

Abstract

Crisis conditions can rapidly reshape collective belief formation by altering the information ecology through which agents observe, share, and interpret risk. We develop an agent-based model of networked social learning to examine how governance regimes—defined by the presence or absence of information suppression, accountability, and hope-based mobilization—reconfigure trust dynamics, rumor reliance, polarization, and action support over an identical exogenous risk cycle. Comparing eight regimes, we identify three robust macrolevel patterns. Open information regimes sustain epistemic coordination, maintaining high average trust, negligible rumor influence, and near-zero horizontal conflict. Suppression without accountability produces an early and persistent transition to rumor-dominant updating, accompanied by trust collapse and durable polarization, yielding an ungovernable equilibrium. Adding accountability mitigates total trust breakdown but does not restore epistemic integration: suppressed-yet-accountable regimes remain stably polarized despite intermediate trust and participation. Hope-based mobilization increases action support across regimes but fails to repair fragmented information environments under suppression, resulting in mobilization without integration. Robustness experiments that relax key assumptions on rumor floors and exposure compression preserve the regime ordering and typology, indicating that the conclusions are structural rather than parameter-specific. These findings highlight how governance configurations condition crisis resilience by reorganizing information ecologies, with implications for the sustainability of coordinated decision-making in organizational and ESG-relevant contexts.