Artificial Intelligence Readiness and Public Service Innovation: A Comparative Malaysia–China Study

ABSTRACT: Artificial intelligence (AI) readiness is increasingly important for public-sector transformation, but the mechanisms through which workforce capability and organizational conditions translate into public service innovation remain underexamined across national settings. This simulation-based comparative study demonstrates an empirical model for Malaysia and China using an assumed sample of 480 public-sector employees (Malaysia n=240; China n=240). Five latent constructs were operationalized: AI training (TR), leadership and innovation culture (LC), employee AI capability (AIC), organizational AI readiness (OR), and public service innovation (PSI). The synthetic measurement model showed acceptable reliability and convergent validity, with Cronbach’s alpha ranging from 0.839 to 0.886, composite reliability from 0.839 to 0.887, and AVE from 0.566 to 0.662. HTMT values were below 0.85. CFA fit was excellent in the simulated dataset (χ²=173.19501898466930356335, df=160, CFI=0.997, TLI=0.997, RMSEA=0.013, SRMR=0.027). Structural results indicated significant effects of AIC on PSI (β=0.216, p<.001), OR on PSI (β=0.424, p<.001), TR on AIC (β=0.457, p<.001), TR on OR (β=0.304, p<.001), and LC on OR (β=0.462, p<.001). The model explained 25.6% of the variance in PSI. Synthetic multi-group analysis suggested that the OR→PSI relationship was stronger in China than in Malaysia (Δβ=0.155, p=0.053), whereas other path differences were not statistically significant. The paper demonstrates a TOE-informed analytical framework suitable for subsequent validation with genuine government-employee data.

KEYWORDS : Artificial Intelligence Readiness; Public Service Innovation; Digital Government; Structural Equation Modelling; Multi-Group Analysis; Malaysia; China.