Lessons from the Field: AI and the Science of Reading

As part of national leadership efforts for the Using Generative Artificial Intelligence for Reading R&D Center (U-GAIN Reading), the listening session AI + The Science of Reading: Accelerate Learning for Multilingual Students convened educators, researchers, district leaders, and students to explore how artificial intelligence (AI) can be thoughtfully integrated into reading instruction for multilingual learners.

The listening session was organized by the U-GAIN Reading National Leadership Cohort and facilitated by Dr. Yenda Prado. Panelists included Lambert Dyshniev, a student at Cabarrus County School District; Gerardo Martinez, executive director of educational technology at Westminster School District; and Carrie Hernandez, teacher at Westminster School District, recognized as Teacher of the Year by the Orange County Department of Education.

Panelist and Participant Perspectives

Panelists and participants shared their experiences with using AI to accelerate literacy development for multilingual students and offered considerations for educators to consider.

How Can AI Support Reading Instruction for Multilingual Learners?

Panelists stressed the importance of instructional alignment. Ms. Hernandez grounded the conversation by clarifying what it means to stay anchored in the Science of Reading. “The Science of Reading is a large body of research that is evidence-based that explains how people learn to read and write, and how to teach them most effectively,” she said. “It emphasizes structured literacy, a systematic, explicit approach to teaching reading and writing.” Ms. Hernandez reinforced that uses of AI must support explicit instruction of foundational reading skills.

Panelists also described how educators are leveraging AI to enhance instruction. Ms. Hernandez shared, “Teachers are using AI platforms to create personalized decodables, add comprehension questions, and tailor reading levels to a passage.” Participants echoed this observation, noting that AI can support planning, differentiation, and feedback for multilingual learners when implemented responsibly. When aligned to evidence-based curricula and goals, AI can help educators better meet the linguistic and academic needs of diverse learners.

Participants also shared their concerns regarding the impact of AI use with multilingual students, including the limited ability of AI tools to hear students correctly and pronounce words correctly. There were also concerns that multilingual speech or accent could be mistaken by the AI as a reading disability.

What Risks and Guardrails Should Educators Consider?

Panelists emphasized that without clear guardrails, AI risks being overused, misunderstood, or misaligned, particularly in classrooms serving multilingual learners. Participant concerns focused on the potential for students’ over-reliance on AI and the impact of AI on students’ brain development, health, and critical thinking. One participant stated, “I worry that young students might start depending too much on AI instead of developing their own critical thinking.”

Dr. Prado highlighted the importance of feedback and monitoring systems to ensure AI tools actually support learning. “Schools must build capacity around feedback systems for ensuring that AI tools are used effectively… monitoring student progress, identifying where instruction is breaking down, and adapting accordingly,” she said. These systems help educators move beyond surface-level metrics and toward meaningful insights about student reading development.

Professional learning also emerged as a critical safeguard. Dr. Prado stressed that educators need support not only in using AI tools, but in critically evaluating them: “Teacher professional learning must include both the Science of Reading as well as digital literacy and AI literacy… teachers need support to understand both how to use AI tools and how to critique them.” Without this dual focus, the use of AI may reinforce inequities or obscure instructional gaps.

Mr. Martinez framed the current moment as an opportunity for deeper alignment. Reflecting on district-level implications, he shared, “AI is forcing us to become aligned pedagogically… to align what we know about literacy with how the brain learns.” Mr. Martinez underscored that AI can catalyze coherence when leaders are intentional about vision, guidance, and shared understanding.

Looking Forward

We closed the listening session by asking, “When you imagine yourself as a leader shaping how AI is used to support student learning, what excites you most?” Panelists and participants were most excited about the possibilities for using AI to assist in personalization and differentiating instruction, with one participant sharing that they were excited about “creating something better than either humans or AI can create alone.”

Mr. Dyshniev shared that he’s most excited about the opportunity to use AI tools to help students learn and build on their ideas. “When I was learning English and programming, I could use AI to easily break down my problems step by step… and explain everything in a way that will make sense to me,” he said. “When I started using AI to help me design new ideas for games and animation, I understood how powerful and cool AI was. A teacher can use it in their class to help students explore topics more deeply and feel more confident learning new things.”

The listening session illuminated the potential of AI use to expand access, personalize learning, and support multilingual readers when it is grounded in learning science and inclusive practice. As U-GAIN Reading continues to convene education and research communities nationwide, this work will require thoughtful leadership, sustained professional learning, clear guardrails, and a shared commitment to ensuring that innovation strengthens teaching and learning.