Classification Drift: Decisions Passing Quietly but Compounding Inequity

Public-sector classification systems accumulate classification drift, a widening gap between what employees actually do, how their work is described, and how they are compensated. Because classification shapes pay, role comparison, and internal equity at scale, classification drift quietly compounds into measurable inequity. At a large U.S. urban public school district (the District), this drift manifested as 435 employees sharing the title “Specialist” across 24 pay grades ($47,000–$125,000). This article describes PRISM (Progressive Refinement and Intelligence Synthesis Model), a multi-model AI framework developed with a research-university partner’s Data Science Institute that introduces validation friction at high-stakes decision points. Rather than automating approvals, PRISM surfaces disagreement between independent AI models and routes ambiguous cases to human reviewers with a structured evidence package. Across 992 positions, 173 cases (17%) were flagged for escalated review. This article describes what public HR offices, with or without AI, can do to address classification drift.