<p>This study investigates how job insecurity affects employees’ cybersecurity behavior by highlighting the mediating role of job stress and the moderating function of self-efficacy in AI use. Drawing on multi-wave data (N = 373) from working adults in South Korea, the research employs a three-time-point design to address longstanding questions about whether perceived threats to employment stability influence compliance with organizational security protocols. Contrary to initial expectations, job insecurity alone did not exert a direct impact on cybersecurity behavior; rather, its detrimental effect was fully conveyed through elevated job stress. These findings indicate that uncertainty surrounding future employment consumes employees’ emotional and cognitive resources, thereby undermining their ability to maintain consistent vigilance in detecting or preventing cyber threats. Additionally, the results reveal that self-efficacy in AI use significantly moderates the link between job insecurity and job stress, suggesting that employees who feel confident in their AI-related competencies experience lower stress under conditions of heightened job insecurity. By synthesizing insights from conservation of resources theory and the job demands-resources model, the study advances both theoretical and empirical understanding of how job-related insecurities and personal technological capabilities collectively shape cybersecurity engagement. The outcomes emphasize the importance of interventions that both alleviate employee stress and enhance skill sets in emerging digital tools, ultimately safeguarding organizational data in fast-changing technological contexts.</p>