Authors: Research Scholar Shefali Sharma, Assistant Professor Dr. Jyoti Kumari
Abstract: Adolescents have become routine users of artificial intelligence (AI) systems faster than school systems have become teachers of them. This paper asks whether AI literacy has the standing of a core twenty-first century competency for secondary education, and what a defensible framework for it would look like. Adopting a structured integrative review, we analysed a purposive corpus of 29 sources across four strands: nine influential competency frameworks, eight review syntheses, six instrument-validation studies and seven system-level survey datasets. Cross-framework mapping shows strong convergence on five constructs understanding AI, using AI, evaluating AI, creating with AI, and positioning oneself toward AI but persistent divergence on whether creation belongs in general education, whether the affective dimension is a competency or a mediator, and how progression should be indexed. Measurement lags conceptualisation: of sixteen validated scales identified in the field, thirteen are self-report and only two are purpose-built and large-sample validated for secondary students, so the higher-order strands are largely evidenced by learners' claims about themselves. System-level data reveal a widening gap: reported use by students rose from 13% to 54% in two survey years, while provision of student training, policy and explicit instruction all remain below 45%. We propose the AILSE framework, which organises the five convergent strands across three progression levels indexed to autonomy, abstraction and accountability, and specifies four enabling conditions teacher capacity, curricular host, assessment validity and equitable access without which the competencies are not attainable. The contribution is a framework that treats implementation not as an afterthought but as part of the construct.
