AI adoption and employee AI skills are linked to stronger competitive advantage in Indonesia's creative sector, but gains depend on a culture of knowledge sharing — technology alone is not enough.

Main Finding

AI adoption, human resource competence, and knowledge-sharing behaviour (KSB) each have significant positive effects on organisational competitive advantage (CA) in Indonesia’s batik and weaving creative industry. KSB also mediates the effects of AI adoption and competence on CA — i.e., AI and competence increase CA partly by increasing knowledge sharing.

Key Points

• Sample: 225 practitioners in the creative industry (batik and weaving) in East Java (purposive sampling).
• Constructs and items:
  • AI adoption (6 items; e.g., use of AI tech for business improvement)
  • Competence (6 items; skills, adaptability, managerial ability)
  • Knowledge-sharing behaviour (5 items)
  • Competitive advantage (5 items)
• Analysis: Partial least squares structural equation modelling (SEM-PLS) using SmartPLS 3.0; bootstrap significance testing (p < 0.05).
• Main hypotheses (all supported):
  • H1a/H1b: AI adoption and competence → positive effect on KSB.
  • H2a/H2b: AI adoption and competence → positive direct effects on CA.
  • H3: KSB → positive effect on CA.
  • H4a/H4b: KSB mediates the effects of AI adoption and competence on CA.
• Model fit / explanatory power:
  • R² for KSB = 0.784 (AI adoption + competence explain ~78.4% of KSB variance).
  • R² for CA = 0.727 (model explains ~72.7% of CA variance).
• Respondent profile highlights: majority female (56.9%), most owners aged 41–50, majority secondary education (56%), many micro firms (52.9% employ 1–4 people), typical operations 5–10 years.
• Limitations noted by authors: cross-sectional design, purposive sampling, self-reported measures, single sector/region.

Data & Methods

• Design: Cross-sectional survey of MSMEs, designers, community managers and cooperatives in batik/weaving centres.
• Sample size: N = 225 (purposive sampling).
• Measurement: 5-point Likert scales adapted from established sources (Pan et al., Braßler & Sprenger, Singh et al., etc.).
• Statistical approach: SEM-PLS (SmartPLS 3.0), measurement model validation (reliability/validity), bootstrapped path significance, effect-size (f²) and predictive relevance reported.
• Ethics: Clearance obtained from Jember University Ethics Committee.

Implications for AI Economics

• Complementarity: The paper provides empirical evidence that AI capital alone is insufficient to generate firm-level competitive gains in resource-constrained creative MSMEs; returns to AI are complementary to human capital (competence) and organisational routines (knowledge-sharing). Economic models of AI adoption should incorporate complementarities between AI investment and knowledge/skill endowments.
• Diffusion and spillovers: High KSB amplifies the impact of AI adoption — implying that policies or platforms that lower knowledge-sharing frictions (peer networks, digital learning hubs, local extension services) can accelerate productive AI diffusion and local spillovers in creative clusters.
• Policy targeting: Subsidies or grants for AI tools should be coupled with investments in training, competence frameworks, and incentives for knowledge-sharing (e.g., collaborative platforms, industry workshops) to increase social returns. For micro firms, low-cost interventions that foster KSB may yield higher marginal returns on AI adoption than hardware/software subsidies alone.
• Measurement and evaluation: Evaluations of AI interventions should measure organisational processes (knowledge flows, routines, skills) in addition to technology uptake and short-term sales; self-reported measures can be complemented by objective performance metrics and longitudinal follow-up to capture persistence and dynamic effects.
• Heterogeneity and scale: The findings suggest heterogeneity in benefits by firm size and capability. Macroeconomic or sectoral projections of AI’s productivity gains should allow for distributional effects — small creative firms may under-realise gains absent complementary investments, affecting aggregate adoption curves and inequality in firm performance.
• Research agenda for AI economics: prioritize causal and longitudinal designs to establish directionality (e.g., does AI induce KSB or do high-KSB firms adopt AI more successfully?), experiment with policy packages (AI tools + training + knowledge platforms), and quantify externalities from KSB-enabled AI adoption across local value chains.