When Inclusion Leads Innovation

Blurring the Lines: What AI Can Learn from Special Education

The most effective strategies in education share a common origin story: they began by designing for the edges.  When educators focused on students with the most complex challenges, they developed tools that proved more flexible, responsive, and powerful for everyone. From differentiated instruction to adaptive technologies, practices once built for access became the blueprint for excellence. hen focused on students with the most complex challenges, they created tools and frameworks that were more flexible, responsive, and powerful for everyone. Many of the practices now considered “best” began as efforts to reach learners once left behind. By creating for those at the margins, we discovered innovations that serve everyone.

For educators, leaders, and developers alike, this is the work ahead: ensuring technology learns from the learners it seeks to serve. 

The future of learning will not be rewritten by technology alone, but by our willingness to design once again for the edges.

When Technology Learns From Us

Artificial Intelligence (AI) may be the next chapter in education’s inclusion story—not because it’s powerful, but because it extends what teachers can see.  AI is already shifting how students access content, how teachers design learning, and how schools define growth—but its purpose will depend on what we choose to value. And that will depend on how we use it.  Our goal is not to replace human connection, but to extend it

The tools emerging today—speech recognition, predictive analytics, and adaptive learning pathways—expand the legacy of inclusion. AI can scaffold reading comprehension in real time, translate content across languages, and provide immediate feedback matched to a student’s pace and mastery level. For a nonverbal student, it might mean finding a voice; for a struggling reader, a new path into meaning. AI transcription tools are helping deaf and hard-of-hearing students participate more fully in discussions; language models can generate leveled reading passages that meet diverse literacy needs in real time.

These are not futuristic dreams—they are extensions of what special education has always done: meet students where they are and build from there.  The individualized supports, differentiated lessons, and adaptive technologies that once lived in special education are now informing how every classroom works. As AI enters this conversation, it carries both the legacy and the responsibility of that work.  With AI, teachers can translate complexity into actionable insights.  The question is not what AI can do, but how we will use it—and whether we will keep students at the center of its design.

Designing for the Edges, Leading from the Margins

The simplest way to understand this principle is to look down at the sidewalk. Curb cuts—those gentle slopes at every street corner—were designed for people in wheelchairs. Today, they are used by everyone: parents with strollers, travelers with luggage, delivery workers with carts. Angela Glover Blackwell called this the “curb-cut effect,” a powerful example of how solutions designed for those most excluded often benefit all of society.

Innovation begins when we design for those most often left out.

In education, the same effect is at work. Frameworks like Universal Design for Learning (UDL) began as efforts to support learners with disabilities but have since become the foundation for inclusive and effective teaching everywhere. UDL focuses on providing multiple means of engagement, representation, and expression—so that all learners can access knowledge and show what they know. By designing for the edges, we create classrooms that are more flexible, equitable, and humane for everyone.

When technology is designed for the margins, it doesn’t just accommodate difference—it redefines what inclusion looks like for all.

Fixing the Wrong Thing

For too long, education has focused on fixing students instead of systems. But what if a student isn’t limited by their learning profile, but by a system’s inability to meet it?  What we see as struggle may be less about a child’s capacity and more about our capacity to respond.

As we build AI tools, this reframing matters. Systems trained on narrow definitions of success risk replicating the same inequities that once made special education necessary. True progress means designing learning environments—and algorithms—that recognize variability as the rule, not the exception. When technology reflects the full spectrum of human learners, inclusion stops being an accommodation and becomes the foundation. That shift—from access to belonging—should guide every decision we make about technology, teaching, and trust. 

Belonging isn’t granted by the system; it’s revealed when the system finally sees every learner clearly. Students aren’t included because we make room for them. They belong because they always did.

The Space Between Independence and Belonging

In education, independence is often seen as the ultimate goal. But independence can sometimes suggest isolation. Callie Turk, co-founder of REEL, asks a question worth sitting with: “Is independence the goal?”

Interdependence may be the better measure—recognizing our shared humanity and mutual reliance. It emphasizes collaboration and belonging, the same principles that define inclusive education.

Yet independence still holds value—not as separation, but as self-determination. When students build the skills and confidence to contribute to the community around them, independence becomes a pathway to interdependence.

Thoughtfully designed AI can help make that possible—not by doing the work for students, but by scaffolding their growth and stepping back as mastery develops. The goal isn’t automation; its agency.  In this way, technology extends—not erases—the circle of support.

AI should build independence by adjusting support dynamically—scaffolding growth without simplifying challenge.

For instance, an AI writing platform might prompt a student to organize their thoughts, but then step back as they gain confidence—echoing the gradual release of responsibility model familiar to educators. A communication app might analyze patterns in speech devices and suggest when to introduce new vocabulary, helping a student express ideas more independently within social interaction, not apart from it. These small, responsive adjustments can reinforce agency while strengthening connection—the essence of inclusive design.

True inclusion—whether through pedagogy or technology—depends on interdependence. When we design for connection, we build systems that see every learner as part of a learning community, not apart from it.

Designing Systems That Support Connection

Until now, AI tools have largely existed in silos. Teachers use them for planning, students for practice, parents for monitoring. In special education especially, those silos are pronounced—IEP-teams, assistive-tech specialists, general classroom teachers and families often operate on parallel tracks rather than shared ones. The next step is AI that bridges those roles—a shared system that builds understanding across the learning ecosystem.

Imagine this: a student’s reading app tracks engagement and comprehension. The AI summarizes for the teacher—“Maya reads best in short bursts.” The parent receives a simpler note—“Maya’s focus improves when reading alternates with movement.” The teacher adjusts, the AI logs it automatically, and everyone stays aligned.

That’s not assistive technology—it’s relational technology, designed to strengthen the human connections that drive learning. If independence defines how learners grow, connection defines how they thrive

Relational technology shifts the purpose of AI from automating tasks to amplifying connection. It recognizes that insight lives in relationships—between teacher and student, between data and context, between learning and life. It helps educators see the learner in full view: not just as a collection of scores or accommodations, but as a dynamic person whose growth depends on understanding across environments. 

Relational technology describes digital systems designed not to automate or isolate, but to connect—linking learners, educators, and families within a shared ecosystem of insight, action, and belonging. In the field of human–computer interaction, scholars such as Carlo Perrotta and Neil Selwyn have described relationality as a framework for understanding how humans and technology co-shape one another. Applied to education, this perspective invites us to see AI not as a tool that replaces professional judgment, but as one that reflects and amplifies it.

As AI grows more practiced, it should not replace human insight but correspond to educator’s understanding of each student, not defining it. In time we could see systems that sync in real time across home and school, flag stress indicators, recommend sensory breaks, and align supports instantly. Not to replace human connection, but to deepen it.

Designing With, Not For

If we want AI to advance equity, we must design it with intention. That means embedding flexibility, transparency, and representation into every layer of development.

Equity isn’t just moral—it’s technical. When datasets underrepresent students with disabilities, systems can misread growth as error or difference as deficit. AI built for the average will fail the margins, but AI built for the margins will serve everyone.

Developers must audit data for representation gaps, collaborate with educators and families, and commit to transparency throughout the design process. As disability rights advocates Michael Masutha and William Rowland of Disabled People South Africa reminded the world, “Nothing about us without us.”  Tools must be built with, not just for, the learners they serve.

Designing for equity means acknowledging that accessibility and ethics are not add-ons—they are the foundation of responsible innovation. Human oversight must remain non-negotiable—AI can inform decisions, but it should never make them.

When Expertise Leads the System

AI is a tool, and in skilled hands, it can help tailor instruction, reveal learning patterns, and free time for what matters most—human connection.

But that depends on trust and training. Educators must trust that AI will amplify their professional judgment, not undermine it; that student data will be protected, not exploited; and that algorithms will clarify complexity, not conceal it. Educators need the time and professional learning to use these tools meaningfully. That trust is built through training—not just in how to operate the tools, but in how to question them, interpret their outputs, and apply them with discernment. Programs like Mursion, which simulates instructional conversations, and adaptive coaching platforms like Aristotal, show what’s possible when AI supports—not supervises—teacher growth.

Trust and training must evolve together. This balance between confidence and competence has long defined effective special education practice—where teachers learn to trust their instincts, question the data, and personalize support.  Educators need both the confidence and the competence to make sound decisions in an AI-supported environment. Special education has long modeled this balance, proving that when pedagogy adapts with both skill and conviction, innovation leads to inclusion.

Special education’s legacy reminds us that lasting innovation begins with inclusion—that the most effective tools are those shaped by the diverse learners they aim to serve. Just as special education once reshaped classrooms for access and belonging, today’s educators must shape AI around equity.

Seeing Every Learner Clearly

Every new tool should be measured against one question: What will this mean for the student?

The legacy of special education reminds us that progress is never about faster systems or smarter tools—it’s about deeper understanding. If AI expands access, strengthens belonging, or deepens that understanding, it honors that legacy. If it reduces learners to data points or replaces relationships with predictions, it betrays it.

As AI becomes more capable, it should not replace human insight but mirror it—reflecting the educator’s understanding of each student, not defining it. 

Building an equitable future for learning isn’t the work of algorithms—it’s the shared responsibility of everyone who teaches, designs, and learns. Technology will keep evolving, but inclusion—and the humanity that drives it—must remain our constant.

November 2025

Winnie O’Leary has spent over 30 years in education as a classroom teacher, school board member, a family advocate, special education teacher, curriculum writer, National Intervention Specialist and Educational Initiative Manager. Winnie earned a BA from Mary Washington College and an MEd from the University of New Orleans. Her experiences have allowed her to work with districts all over the country where she finds something new and exciting every day.