Answers are grounded in this paper and cite their sources.
Digital inclusion with generative AI depends not only on educators’ access to tools but on how they use them. Comparing the benefits educators perceive with the uses they report may reveal gaps between what educators believe AI can do and how they use it in teaching. In this exploratory secondary analysis of survey data from 293 Minnesota PK-12 and higher education educators, we compared endorsement of improved accessibility for diverse learners as a benefit of generative AI with reported use of AI to enhance accessibility for students with disabilities. Because the items differ in scope and target population, the comparison describes response proportions, not a failure to act on a belief. Among 160 AI-using educators who answered both items, 133 (83.1%) endorsed the benefit and 51 (31.9%) reported the use (McNemar exact p < 0.001). Subgroup differences were not detected, though power was limited. Among all AI users, self-rated competence was associated with reported accessibility use (44.0% vs. 27.9%; adjusted OR = 2.31, 95% CI [1.13, 4.73]); among endorsers, the estimate was imprecise. Keyword-selected open-ended responses more often described general differentiation and language support than disability-specific use. We discuss measurement implications and propose supports as hypotheses for evaluation.
No takes yet. Share an insight, caveat, or question.
Martinez et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: