AI Breaks the Proxy: A Framework for Transforming IS Pedagogy from Execution to Judgment

Generative AI has broken the execution proxies that IS education has relied on for decades. When students can produce correct SQL queries, functional web applications, and polished dashboards in minutes, execution no longer serves as evidence of understanding. This paper introduces the Artifact-Judgment Model (AJM), an assessment framework that reverses the traditional relationship: the artifact is given, the judgment is assessed. Students receive AI-generated work products and demonstrate understanding through evaluation, prediction, and trade-off reasoning rather than execution. The AJM is AI-proof by design — because assessment starts from AI output, using AI to complete it would simply reproduce the errors the assessment is designed to catch. The model is grounded in a three-tier framework analyzing how AI transforms IS competencies and is applied across six core IS domains. It aligns with AACSB 2026 Standard 4.3’s mandate for human judgment over tool mastery and ABET’s emphasis on AI as a support for, not replacement of, human judgment. The proxy problem is not unique to IS; the AJM offers a generalizable response to the defining pedagogical challenge of the AI era.