IEPs in the Age of AI: Examining IEP Goals Written with and Without ChatGPT
Danielle A. Waterfield, Olivia Fudge Coleman, Nathan P. Welker, Michael J. Kennedy, Sean D. McDonald, Bryan G. Cook
Abstract
Individualized Education Programs (IEPs) are a core element of special education in the United States. Within IEPs, IEP goals drive the implementation of IEPs and guide measurement of progress for students with disabilities. Yet research indicates that many IEP goals lack sufficient detail, indicating overall low-quality goals. Additionally, special education teachers can feel unsupported in their jobs and struggle with managing their workloads. This convergent mixed methods study explores the integration of artificial intelligence (AI) in special education to address these issues. Specifically, we explored how experienced teachers perceive AI’s role in their practice and compared the quality of AI-generated IEP goals to those written by special education teachers. Quantitative findings show no statistically significant difference ( p = .67) in quality ratings of IEP goals written only by teachers versus AI-generated goals. Qualitative findings depict overall positive perceptions on using AI to facilitate workload. Implications and opportunities for future research and the field centering on continued exploration and training of using generative AI to assist special education teachers are discussed.