Learning after COVID-19 and the ICT career aspirations: Are students entering the AI era with weaker skills?

Synthesis: Post-COVID ICT Career Aspirations uses PISA 2018 and 2022 country-level data to investigate whether students entering the generative AI era have adequate educational foundations. Using a mixed-methods approach including Variational Autoencoders for latent representation learning, the study finds that ICT career aspirations have increased globally but unevenly. Digital skills are the strongest and most consistent predictor of ICT aspirations, while teacher support plays a complementary role and student autonomy shows only weak, context-dependent effects. The findings challenge simplistic narratives about AI Literacy — rising interest in ICT careers does not automatically mean students have the foundational skills to succeed in AI-driven labor markets. This connects to The Illusion of Competence: Self-Perceived Digital Literacy and AI Readiness Among European Secondary Students findings about student overconfidence, and to AI Adoption Among Teachers: Insights on Concerns, Support, Confidence, and Attitudes research on the complementary role of teacher support. The multidimensional nature of educational readiness aligns with Teacher AI Competency frameworks, and the uneven global distribution of digital skills raises Equity concerns previously documented in .

What this means for practice

• Instructors. Build students' digital skills deliberately instead of assuming that exposure to technology will raise their interest in ICT careers — digital skills were the only consistently positive predictor of aspiration growth (β = 0.072).
• Instructors. Pair any move toward greater student autonomy with visible instructional support; in the discriminant analysis teacher support ranked just behind digital skills among high-growth systems, and autonomy alone carried a negative standardized coefficient (β = -0.192).
• Administrators. Fund pedagogical and institutional support alongside infrastructure: the analysis concludes that digital infrastructure investments are unlikely to produce sustained increases in ICT career interest without complementary teacher support and curricular integration.
• Learners. Treat interest in an ICT career as a starting point that has to be backed by demonstrated digital competence, since aspiration grew unevenly across countries while competence remained the strongest correlate.

Limitations

• This is a country-level secondary analysis of PISA 2018 and 2022: student autonomy, digital skills, and teacher support were aggregated at country level, so no claim about an individual student can be supported, and only systems with complete data across both waves entered the sample.
• The learning-environment indices come from PISA 2022 while the aspiration changes span 2018–2022, and none of the measures captures generative-AI skill directly.
• The regression puts teacher support (β = -0.324) and autonomy (β = -0.192) in a negative association with ICT aspiration growth against a small positive digital-skills effect (β = 0.072), which the authors attribute to complex, context-dependent dynamics rather than a clean causal story.
• The discriminant model's classification accuracy was moderate and the high- and low-growth groups overlapped substantially, so learning-environment indicators do not fully determine ICT career trajectories, and labor market and socio-cultural factors are left unmodeled.

Citation

Diana Maria Popa, Simona-Vasilica Oprea, Adela Bâra (2026). Learning after COVID-19 and the ICT career aspirations: Are students entering the AI era with weaker skills?. arXiv preprint.