Procedural Legitimacy Without Absorption: AI, Functional Illiteracy, Credential Inflation, and the Institutional Misalignment of U.S. Higher Education

This paper presents a systems-level analysis of contemporary higher education, examining why long-standing concerns around assessment validity, credential inflation, labor-market underemployment, and the governance of generative AI persist despite sustained reform efforts. Rather than treating these issues as isolated failures or as consequences of individual behavior or technological change, the paper advances a unified diagnosis centered on procedural legitimacy—the allocation of recognition through standardized, auditable processes that are increasingly decoupled from demonstrable competence and economic absorption. Drawing on literature from higher-education governance, labor economics, literacy research, and science and technology studies, the analysis develops a closed-loop legitimacy model in which institutional stability is maintained even as epistemic reliability and labor-market alignment degrade. Generative AI is examined not as a primary cause of dysfunction, but as an accelerant that exposes the fragility of artifact-based assessment and intensifies existing misalignments under conditions of informational abundance and heterogeneous cognitive capacity. The paper further demonstrates how population-level literacy constraints shape the effectiveness of AI-mediated learning, how credential inflation weakens signaling value, and why underemployment among highly credentialed individuals functions as an outcome-level indicator of systemic decoupling rather than individual failure. It argues that responses focused on individual adaptation, compliance, or tool restriction are insufficient because the failure mode is institutional. As treatment, the paper outlines a framework for institutional innovation that rebalances legitimacy allocation without abandoning procedural governance. Proposed directions include outcome-sensitive competence verification, AI-inclusive assessment designs that restore cognitive traceability, first-class competency-based recognition pathways, and feedback mechanisms linking credential production to labor-market absorption. Historical and international precedents are discussed to demonstrate feasibility without prescribing a single reform model. The work is intended for interdisciplinary academic, policy, and institutional audiences concerned with the future governance of knowledge, education, and expertise under contemporary economic and technological conditions. The paper is released under a CC-BY-SA 4.0 license to encourage open critique, reuse, and extension.