Attitudinal openness to AI as supervisor or colleague/subordinate in Japan: the role of advanced digital technology familiarity.

Organizations adopting artificial intelligence (AI) must account for employees' attitudes toward AI in different workplace roles. This study examines whether familiarity with advanced digital technologies is associated with openness to AI as a supervisor or as a colleague/subordinate in Japan. We analyze a three-wave balanced panel of continuously employed respondents using hierarchical Bayesian panel models. The focal measure is the Advanced Digital Technology Familiarity Index (ADTFI), based on self-reported familiarity with six technologies; the outcome is stated openness to hypothetical AI role arrangements rather than realized adoption. Higher ADTFI is positively but modestly associated with workplace AI role openness. A covariate-adjusted coefficient decomposition shows that computing capability and workplace technology exposure have positive ADTFI-related components, while workplace technology exposure has an imprecisely estimated negative conditional component and a combined component whose interval includes zero. Role-specific ordered-probit models yield a negative TechAdop coefficient with an interval excluding zero for the AI-supervisor item, whereas the corresponding colleague/subordinate estimate is near zero and imprecise; this descriptive contrast is not a formal cross-role coefficient test. A lagged model retains a small positive association between prior ADTFI and subsequent openness after prior openness is controlled. The results mainly reflect stable between-person differences and are conditional on continuously employed panel respondents. They should therefore be interpreted as descriptive association patterns, not causal effects of training or AI implementation.