| Abstract: |
Supported by an Institute of Education Sciences grant, the U-GAIN Reading R&D Center is investigating artificial intelligence (AI) applications aligned with reading science to achieve equitable gains for diverse elementary students, including English Learners (ELs). This report explores how generative AI (GenAI) can be used by schools to enhance reading instruction, particularly for ELs, in light of declining national literacy rates. The study outlined in this report involved a market scan of edtech products, interviews, and listening sessions with 16 educators across 11 states. Findings indicate that while many districts prioritize the Science of Reading, implementation varies. The market scan revealed that platforms are increasingly incorporating elements that reflect the whole child based on the Science of Reading. This includes the incorporation of social-emotional learning elements into stories, engaging 1:1 tutoring opportunities, and real-time assessment and micro-interventions. Educators interviewed indicated that they are integrating AI tools primarily for personalized learning and language support. They also indicated frequently using generative AI tools for lesson planning. Key opportunities discussed include uses of AI to customize texts to student interests and reading levels, engage learners in adaptive listening and speaking activities, and foster natural discussion through the use of culturally responsive content, particularly for ELs. Significant concerns were also raised regarding biases and stereotypes in AI-generated content and images, inaccuracies in Automated Speech Recognition (ASR) for diverse accents, cultural biases, and misuses of student data privacy. The report concludes with recommendations for designing AI tools with ELs at the forefront within the Science of Reading context. This includes ensuring transparency in AI development, as well as establishing robust district policies for vetting AI tools to ensure ethical and culturally responsive adoption. |