Unveiling AI data security: How employee awareness evolves in smart manufacturing

Abstract

In the era of Industry 4.0, smart manufacturing leverages artificial intelligence (AI) to enhance operational efficiency, yet heightened data security risks underscore the critical role of employee data security awareness (DSA). This study pioneers a Cellular Automata (CA) model, grounded in Social Cognitive Theory (SCT), to investigate the emergent dynamics of employee AI DSA in smart manufacturing enterprises. By integrating local security climates and dynamic threshold mechanisms, the model simulates collective awareness evolution under three scenarios: no intervention, mild publicity, and mandatory training, using an initial distribution of 30% low, 40% intermediate, and 30% high-awareness employees. Findings reveal that without intervention, awareness fluctuates unstably, with low-awareness employees rising to 50% and high-awareness declining to 20%, driven by intermediate-state volatility. Mild publicity boosts high-awareness to 45% and reduces low-awareness to 25% (13.3% overall increase), while mandatory training elevates high-awareness to nearly 80% and suppresses low-awareness below 5% (37.8% overall increase). Sensitivity analysis validates model robustness, highlighting intermediate-state employees as pivotal drivers of awareness dynamics. This study advances SCT by quantifying triadic interactions in AI-driven contexts and offers actionable insights for optimizing data security through targeted interventions, demonstrating that hybrid strategies combining publicity and training yield superior outcomes.