Beyond Tool Adoption: A Practical Five-Stage Developmental Continuum for AI Literacy in Higher Education

URL SCAN: Beyond Tool Adoption: A Practical Five-Stage Developmental Continuum for AI Literacy in Higher Education
FIRST LINE: Artificial intelligence (AI) literacy is increasingly recognized as a foundational competency for all university graduates.

I. THE DISSECTION

This paper performs the function of institutional transition management — it acknowledges AI displacement anxiety and provides a framework for treating the symptom (student confusion about AI tools) while leaving the underlying disease (structural labor market collapse) entirely unaddressed. The five-stage model — Not Yet Engaged → Uncritical Use → Informed Use → Critical Evaluation → Improvement — is a pedagogical scaffolding designed to make human engagement with AI tools more orderly. It is, at core, a curriculum design document dressed in the language of developmental psychology.

The framing is relentlessly upbeat. Students are "engaging" with AI. Courses produce "evidence of movement toward informed engagement." The language of "foundational competency" and "responsible application" implies that the purpose of this intervention is to produce graduates who can participate meaningfully in an economy that still needs them.

The hidden assumption: This entire paper assumes that human AI literacy is a survival variable — that developing better frameworks for teaching students to use AI tools will yield economic or social benefits for those students. The DT framework rejects this premise categorically. AI literacy as described here is not a survival asset. It is at best a lag-extending comfort item — it helps institutions manage the transition while doing nothing to prevent the structural outcome. It is teaching better penmanship to people whose jobs are being eliminated by word processors.

II. THE CORE FALLACY

The Fallacy of Productive Participation Repair

The paper's foundational error is assuming that "critical engagement with AI systems in disciplinary and societal contexts" constitutes a viable economic position. It does not. Under the DT framework, the question is not whether students can evaluate or responsibly apply AI. The question is whether their cognitive labor has durable economic value independent of AI systems. The continuum describes behaviors — using tools, evaluating outputs, improving prompts — that are precisely the tasks most susceptible to AI replacement. Teaching students to be better users of systems that are replacing their capacity to be useful is not education. It is grooming.

The secondary fallacy is methodological. The paper admits it has no validated pre/post instrument and no comparison group. The findings are "observational and practice-based." This means the entire evidence base is self-referential: participants moved toward "behaviors consistent with" the framework because the framework defined the behaviors being observed. This is not research. This is program evaluation dressed as scholarship.

III. SOCIAL FUNCTION

Classification: Transition Management / Elite Self-Exoneration

This paper serves the institutional function of making universities look like they are responding actively to technological disruption, without actually confronting what that disruption means. NC State's 330 participants and "credit-bearing courses" create the appearance of meaningful intervention. The UNESCO and OECD alignment elevates the paper's prestige and provides cover: international bodies endorse AI literacy, therefore this work is consequential.

What it actually does is give administrators a framework to show legislators and accreditors — look, we are preparing students for the AI economy. Meanwhile, the preparation being offered is precisely calibrated to the tasks that will be automated first. The authors know this is weak. They hedge with phrases like "observational and practice-based" and "we frame the findings as." But the framing of the problem — "students cluster at two problematic extremes: avoidance and uncritical reliance" — reveals the core anxiety: institutions are terrified of students who either reject AI (they won't be productive) or embrace it uncritically (they won't be controllable). The solution is to produce the right kind of compliant engagement.

This is not education. It is behavioral conditioning for an economy that will not need most of the students it is being delivered to.

IV. THE VERDICT

The paper is professionally constructed, methodologically honest about its limitations, and substantively useless as a structural intervention. It describes a well-designed pedagogical framework for producing better AI tool users within a system that does not need better AI tool users — it needs fewer of them. The five-stage continuum is a refinement of a training paradigm for obsolescence. "Critical evaluation" and "improvement-oriented practice" in the context of AI tool use describe precisely the workflows that AI itself will optimize better than any human trained in this framework.

The case study from NC State — 330 participants, Fall 2024–Spring 2026 — is too small, too observational, and too self-referencing to constitute evidence of anything except institutional activity.

Survival viability of the framework's graduates: Fragile at 1 year, Terminal at 5 years. The skills being taught are precisely the inputs to AI systems, not the outputs that AI systems cannot replicate. A student who becomes "informed" about AI tool use is a student who has invested significant time and institutional resources into developing competencies that will be automated before they enter the workforce at scale.

What this paper actually proves: Universities will continue to produce transition-management documents that look like responses to technological displacement while structurally failing to address what displacement means for the humans they are supposedly educating.

V. AXIOM COMPLIANCE

The paper explicitly violates Axiom 2 (Replacement, Not Survival): It frames AI literacy as a pathway to productive economic participation when the DT framework holds that productive participation requires structural economic necessity — not tool operation competence. The authors do not engage with the possibility that AI literacy as defined may be a signal of obsolescence acceleration rather than protection against it. The paper is an institutional document, not a structural analysis.