Work samples are one of the most powerful and overlooked sources of learning data in special education. This session shows how AI can support, not replace, professional judgment when analyzing student writing, math, and classroom tasks. Participants will learn practical, research-informed procedures for identifying error patterns, task demands, skill gaps, and instructional mismatches using authentic student work. We will also address ethical use of AI, including privacy, bias, transparency, and maintaining evaluator responsibility in decision-making. Through guided examples, you will see how thoughtful work-sample analysis can go beyond surface-level mistakes to yield clearer, more defensible impact statements that connect student performance to standards, classroom expectations, and eligibility decisions. You will leave with practical tools you can use immediately in your evaluations.