Longitudinal Impact of AI Interventions on Skill Development and Employability

Abstract

This study investigates the long-term effects of AI-supported interventions on learners’ skill development and future employability across different social and educational settings. Using a multi-site, mixed-methods approach, 2,400 participants aged 18–35 from vocational schools and higher education programs were tracked over 24 months. Participants received one of three AI-enabled interventions—adaptive learning platforms, AI-powered resume and career coaching, or simulated work environments—while comparable groups received traditional instruction. Quantitative measures included domain-specific skill acquisition (assessed through validated competency tests), overall cognitive and metacognitive improvements, employment status, job quality (wages, contract types), and job stability; the analysis employed latent growth modeling and propensity score-adjusted survival analysis. Qualitative interviews and employer surveys explored perceptions of skill relevance and workplace integration. Findings reveal that AI-supported adaptive learning led to significantly faster initial skill gains (Cohen’s d = 0.45) and better skill retention at 12 and 24 months, whereas AI career coaching enhanced job-search effectiveness and reduced time to employment by 27% relative to controls. Simulated work environments were most strongly correlated with employer-rated job performance, particularly on technical and problem-solving tasks. Moderation analyses show that baseline digital literacy and institutional support considerably influenced effect sizes. The study highlights mechanisms by which AI interventions improve both measurable skills and labor-market outcomes while also addressing equity considerations. Policy and practice implications include creating integrated AI-human curricula, investing in digital literacy support, and tailoring AI approaches to local labor market needs.