This study aims to analyse how the explainability and perceived autonomy (PA) of agentic artificial intelligence (AAI) influence employees’ intentions to use such systems. Recognising that AAI systems differ from previous technologies, this study explored the mediating and moderating roles of knowledge-sharing culture (KSC) and technical self-efficacy (TSE), respectively. The objective was to understand the cognitive, cultural and individual contexts in which employees would like to use AAI.
We used structural equation modelling to analyse data collected from 719 employees in different British companies.
Results reveal that explainability is a robust predictor of intention to use AAI, exceeding the impact of PA. PA was found to significantly influence KSC, a social and cognitive factor that shapes AI traits and drives adoption intentions. Findings also showed that TSE enhanced the mediating effect of explainability while not affecting autonomy, underscoring the unique psychological mechanisms governing human–AI interactions.
Organisations should prioritise explainability and invest in systems that encourage employee knowledge-sharing and sense-making, including AI discussion platforms and tools that promote transparency. Adjusting AAI autonomy and boosting employees’ TSE can encourage adoption, particularly when AI reasoning must be explained.
This study extends existing AI adoption and knowledge management perspectives by examining how KSC helps translate explainability and PA into employee adoption intentions. It emphasises explainability as a predictor and identifies KSC as the crucial link connecting AI attributes to employee intentions.
