Answers are grounded in this paper and cite their sources.
Purpose This study examines how employee-perceived generative artificial intelligence (GenAI) governance maturity relates to approved enterprise use and unauthorized public-tool use in data-sensitive work. Design/methodology/approach A two-wave survey of 750 knowledge workers in Taiwan was analyzed using partial least squares structural equation modeling (PLS-SEM). Additional checks examine same-occasion measurement, available controls, sector sensitivity and an alternative assimilation-to-shadow-use path. Findings regulatory pressure and internal guideline salience were positively associated with governance maturity. Governance maturity was positively related to safe-to-use climate and perceived data exposure risk. The former was associated with sanctioned assimilation depth and task performance; the latter was negatively associated with shadow GenAI Use, which was positively related to employee creativity. These associations suggest a tension between reliable sanctioned performance and risk-laden exploration. Originality/value Sanctioned assimilation depth and shadow GenAI use are distinct appropriation modes, not opposite ends of one use continuum. Integrating institutional and socio-technical perspectives explains how governance differentiates the conditions attached to approved and unapproved channels without assuming that consumer-led workplace technology use is new.
No takes yet. Share an insight, caveat, or question.
Stanley Y. B. Huang (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: