The Illusion of Competence: Self-Perceived Digital Literacy and AI Readiness Among European Secondary Students

Synthesis: This multicenter study (N=243 European secondary students) systematically challenges the 'Digital Native' paradigm by demonstrating a severe confidence-competence gap in digital and AI literacy. Students report near-maximum Self-Efficacy in passive digital consumption (browsing, social media) but exhibit a sharp decline when evaluated on active technological creation and algorithmic logic — a collective Dunning-Kruger effect.

The AI Paradox

A particularly striking finding is the 'AI Paradox': students significantly overestimate their critical awareness of deepfakes and algorithmic biases compared to their operational AI skills. This creates a false sense of invulnerability against modern misinformation — students believe they can detect fakes but lack the practical skills to do so. This directly connects to emerging research on cognitive offloading and AI-induced miscalibration.

Context-Dependent Gender Gap

The study finds that the technology gender gap is not universal — it emerges significantly only within Technology-oriented classrooms (p=0.046), implicating stereotype threat in formal STEM environments. This has implications for how AI literacy programs should be designed and targeted, particularly in K-12 settings where self-selection into technology tracks amplifies pre-existing disparities.

Pedagogical Imperative

With 76.5% of students demanding pedagogical reform toward hands-on, active technological creation, this work aligns with calls to move beyond passive theoretical instruction. It supports the ECD-based governance frameworks that emphasize observable student competencies over abstract knowledge testing. The findings also connect to broader questions about AI fatigue — students who think they know more than they do may paradoxically disengage from learning opportunities.

What this means for practice

• Learners. Benchmark your digital skills on real tasks before trusting your self-rating: students here scored near the ceiling on turning on a computer (M = 4.88) and managing privacy (M = 4.81) but dropped to M = 2.86 for CAD and M = 2.96 for programming.
• Learners. Treat critical AI awareness as unproven until you can demonstrate it operationally — students rated their ability to spot biases and deepfakes (M = 4.28) higher than their prompting skills (M = 4.15, p = 0.015).
• Learners. Choose coursework and electives that require active creation, not passive use: Technology-pathway students reported the highest active-creation self-efficacy (M = 3.43) against Arts and Humanities students (M = 3.04, p < 0.001), and device exposure alone did not close the gap.
• Learners. Push for the hands-on reform the respondents themselves asked for — 76.5% demanded a shift away from theoretical instruction toward practical technological work.
• Learners. Do not read a technology-track environment as a judgment on your ability: the gender gap in active-creation confidence appeared only within the Technology pathway (n = 111, p = 0.046) and was absent in Sciences (p = 0.142) and Arts and Humanities (p = 0.841).

Limitations

• The sample is a convenience sample of 243 students aged 12 to 16 drawn from just three institutions — one in Valladolid (Spain), one in Zamora (Spain), and one in County Westmeath (Ireland) — and the authors caution against over-generalizing.
• Every measure is self-reported on a 5-point Likert scale, so the confidence–competence gap is a gap in perceived competence and cannot be read as proof of actual skill or its absence.
• The gender comparison rests on n = 241: two non-binary respondents (n = 2 of N = 243) were excluded from the comparative demographic analysis to eliminate any de-anonymization risk.
• The design is cross-sectional with no intervention or follow-up, so the pedagogical reform the respondents demand is not tested for effectiveness here.

Citation

Rodriguez-Alvarez, N., Blanch-Marsolini, A. M., Vara-Gutierrez, S., Gil-Garcia, H., Calzon-Dueñas, J., & Rodriguez-Merino, F. (2026). The illusion of competence: Self-perceived digital literacy and AI readiness among European secondary students.