AI Won’t Transform Teaching and Learning — Unless It’s Built for All Learners

Posted on • Reading Time: 4 min read

This article was written by Erin Stark and published by New Schools on November 19th, 2025.

Our education system was never built with every learner in mind, especially students who learn and think differently. One in five students has a learning difference, yet most classrooms and tools still aren’t designed for them, limiting their ability to thrive. As artificial intelligence (AI) begins to reshape teaching and learning, we risk repeating the same inequities the system was built on. But, we have an opportunity in front of us — we can design AI solutions to meet the needs of diverse learners from the start, moving us closer to an education system that truly delivers opportunity for all.

Teachers shoulder the challenge of meeting the diverse needs of their students every day. Many spend hours adapting or creating instructional materials on their own, often without adequate support. It’s no surprise many feel unprepared and that a widespread belief persists among educators that students with learning differences can’t meet grade-level standards. Yet research shows 90 percent can with the right support. The problem isn’t the teachers or the students. It’s the barriers woven into the system itself — curricula that aren’t designed for all learners, tools that don’t talk to one another, and schedules that leave little room for collaboration.

Past innovations like text-to-speech, audiobooks, and closed captioning expanded access for people with disabilities and improved learning for everyone. AI offers another opportunity to advance inclusion when it’s designed with diverse learners in mind. Students with disabilities are not a monolith. Their strengths and needs vary widely, and no single approach works. That complexity makes it difficult for teachers to meet every need,  but it’s precisely where well-designed AI can make a difference.

Here are three ways funders and innovators can ensure AI expands opportunity instead of repeating past inequities.

Build AI that Understands What Works for All Learners

Most large and small language models can describe instructional practices that help students with learning differences, like explicit instruction or scaffolding. But they rarely model how or when to use them. There is an opportunity to change that by building open, research-based datasets that reflect how diverse learners learn and how effective teachers teach.

AI should be trained on three interconnected instructional practices that work in concert — from accessibility standards like the Web Content Accessibility Guidelines, to Universal Design for Learning, to evidence-based strategies such as structured literacy for dyslexia and visual representations for dyscalculia. Imagine a system trained on that foundation, able to recommend and design instruction with the same nuance as a seasoned special educator.

Open, research-based datasets built on this foundation could power training methods that help AI learn when and how to apply the right supports.

Use AI to Increase Access to Grade-Level Instruction

See demo videos

Efforts to personalize learning can often lower expectations. When students struggle, content is simplified rather than scaffolded, leaving those who need the most support with the least access to rigorous material. For students with learning differences, that pattern is especially harmful — not because they can’t meet grade-level standards, but because the tools meant to support them make those standards less accessible.

AI can help reverse that trend. Tools designed with high expectations and the right support can keep students engaged in rigorous content that enables them to achieve grade-level mastery. For example, literacy tools can make texts more accessible, reformatting passages for easier readability or reading them aloud without lowering text complexity. Math tools can provide just-in-time support, using visuals such as number lines or blocks to reinforce key concepts when students get stuck. And teacher tools can analyze student work or assessment data to flag which skills need attention next and suggest evidence-based strategies to address them. The goal is a shift from simplify to support toward scaffold to succeed — empowering teachers with insights that preserve rigor while meeting every learner where they are.

Break Down Silos Between General and Special Education

See demo video

Teachers navigate a maze of disconnected systems — student information platforms, IEP databases, assessment tools, and lesson-planning software that rarely communicate with one another. The result is fragmented instruction and inconsistent support, especially for students whose growth depends on having coherent experiences across settings. General and special educators often work in parallel rather than in partnership, with critical insights about student progress, accommodations, and goals buried in separate systems. This disconnection not only limits instructional alignment but also adds hours of administrative burden for teachers already stretched thin.

AI can help unify what has long been divided. Tools that integrate IEP goals, progress monitoring data, and classroom instruction can create a more connected ecosystem where educators share a clear view of each student’s learning journey. These tools can automate documentation, surface actionable insights, and make it easier for teams to coordinate support in real time. When a special-education and classroom teacher can instantly see the same data, they gain time for planning and for relationship building with students that drive learning. The outcome is more coherent instruction, stronger collaboration, and more time for what matters most — teaching.

A New Vision for Inclusive Innovation

The future of education AI shouldn’t just be about efficiency or personalization. It should also be about equity and coherence. When models are trained on inclusive, evidence-based data; when they scaffold rather than simplify; and when they connect teachers across traditional divides, AI can serve as a force for systemic change. It can bring educators together and ensure that innovation and inclusion move together, so every student can flourish.