Paper Type
ERF
Abstract
As e-government systems increasingly adopt artificial intelligence (AI), questions of public trust, transparency, and privacy become critical barriers to—or enablers of—e-participation and e-inclusion. This emergent research investigates what conditions shape citizen trust in AI-driven digital government services, and how trust—or its absence—shapes the character of citizen engagement. Grounded in the Technology Acceptance Model (TAM) and institutional trust theory, three themes emerged from semi-structured interviews with 35 participants: trust is source-dependent, transparency is a prerequisite for meaningful consent, and privacy fears constitute a structural barrier to e-inclusion. Findings suggest that distrust does not produce disengagement—it produces qualitatively diminished engagement: more guarded, resistive, and less authentic. This research identifies an opt-out paradox: in compulsory service contexts, citizens must engage regardless of trust disposition, shifting the relevant outcome from willingness to engage toward the character of engagement. Implications for e-participation design and AI governance are discussed.
Paper Number
1335
Recommended Citation
Gills, Kelvis C. and Oluwalowo, Roxanne, "Compelled But Not Cooperative: Public Trust and Engagement Character in AI-Enabled Government" (2026). AMCIS 2026 Proceedings. 3.
https://aisel.aisnet.org/amcis2026/egov/sig_egov/3
Compelled But Not Cooperative: Public Trust and Engagement Character in AI-Enabled Government
As e-government systems increasingly adopt artificial intelligence (AI), questions of public trust, transparency, and privacy become critical barriers to—or enablers of—e-participation and e-inclusion. This emergent research investigates what conditions shape citizen trust in AI-driven digital government services, and how trust—or its absence—shapes the character of citizen engagement. Grounded in the Technology Acceptance Model (TAM) and institutional trust theory, three themes emerged from semi-structured interviews with 35 participants: trust is source-dependent, transparency is a prerequisite for meaningful consent, and privacy fears constitute a structural barrier to e-inclusion. Findings suggest that distrust does not produce disengagement—it produces qualitatively diminished engagement: more guarded, resistive, and less authentic. This research identifies an opt-out paradox: in compulsory service contexts, citizens must engage regardless of trust disposition, shifting the relevant outcome from willingness to engage toward the character of engagement. Implications for e-participation design and AI governance are discussed.
Comments
SIG E-GOV