“So Much to Say, No Way to Share It”: How AI Can Change the Game for Learners with Disabilities

For Jillian, a freshman in high school, her every movement and act of communication has to be intentional, thought out, and often translated through someone else’s interpretation. Born with cerebral palsy and diagnosed with autism, she is non-verbal, requires a wheelchair, needs a GI tube for eating, and has little control over her body. 

Regardless of her physical barriers, Jillian has always been curious, drawn to how technology can assist her. At a young age, she started using an eye transfer board, a low-tech, non-electronic communication tool that allowed her to provide basic “yes” or “no” answers. When Jillian was eight, she discovered the Tobii eye-tracking device, which allows individuals to control a computer and communicate using only their eyes. “The hardest part,” her dad Mick says, “was knowing she had so much to say, but no way to share it.”

After six years of evaluations and insurance approvals, Jillian finally received her Tobii and her world began to shift. The day she realized she could change a streaming channel on her own, she understood she had the autonomy to change more. At the hospital, she used her device to tell nurses she felt sick and needed to lie down, or to advocate for herself against being given unnecessary medication. “That was profound,” Mick remembers. “She could finally tell us what her body needed.”

“The hardest part was knowing she had so much to say, but no way to share it.”

Some teachers didn’t believe Jillian could read or do math until they saw her using her Tobii to provide comprehensive responses and solve math problems. However, challenges persist. Eye-tracking calibration to set up the device can be physically exhausting. Staff turnover means constantly retraining people to understand her communication style. And, well-meaning adults sometimes finish Jillian’s sentences before they give her the chance to respond. 

The AI Game Changer

Technology has long helped students with diverse educational needs—but artificial intelligence (AI) is a game changer, according to Bruce Alter, a physical therapist and assistive technology consultant for the Oregon school district Jillian attends. “This is the first time we’ve had the tools that are needed for the most severely disabled kids to fully express themselves and to achieve and to have high academic achievement.”

Alter, who has learning disabilities, has spent years working and designing for students whose intelligence and curiosity outpace the tools they’d been given to express themselves. Many students with severe disabilities rely on augmentative and alternative communication tools and strategies that support or replace speech. But the real problem, Alter said, is getting these students literally plugged into the classroom—for example, supporting them to take notes in class via keyboard or join classroom discussions. 

“Imagine trying to write a paragraph while half your attention is on how to physically get each word out, and the other half is desperately trying to hold onto the thought you started with,” Alter explained. That’s where AI is beginning to change everything. He built a system that transcribes a few spoken keywords and, with a single keystroke or eye-gaze selection, uses a large language model to expand them into a full response. By building a core vocabulary tailored to each learner and adding predictive text, AI can suggest words and phrases in sequence, speeding communication without sacrificing voice or choice. For the first time, students could write longer, more complex thoughts and then iterate and refine them, like their peers. AI can also continue adapting to students’ communication style as they rotate across school staff members, giving them more autonomy and independence while helping educators. 

“This is the first time we’ve had the tools that are needed for the most severely disabled kids to fully express themselves and to achieve.”

The impact has been profound. Harper, a bright third grader with cerebral palsy, was long underestimated, her wheelchair and soft-spoken manner often leading people to overlook her academic ability. Using AI, she scripted and delivered a goodbye speech to her class, a project that showcased her wit and intelligence. For Alter, it was a glimpse of what happens when you remove the communication barrier and raise expectations.

AI is also opening doors beyond speech: For visually impaired students, smart phones can capture text from a photo and read it aloud. AI can generate alt text to describe charts and graphs through audio, customize reading materials for grade level, and rapidly tailor assignments to provide students with materials that match their interests and abilities. 

Evidence-Based Inclusive Design

A growing body of research and practice supports integrating edtech and AI into education for learners with disabilities if the tools are designed and implemented responsibly. For example, Rapid Online Assessment of Reading, an AI-supported research-based project of the Stanford Accelerator for Learning, created an online platform for assessing early signs of reading difficulties and dyslexia that is now a screening tool in California. The Autism Glass Project at Stanford developed smart glasses that use machine learning to improve children’s ability to recognize and respond to social cues. In such cases, Isabelle Hau, executive director of the Stanford Accelerator for Learning, emphasizes the importance of co-designing with people with disabilities and placing them in the driver’s seat of innovation.

Similarly Jutta Trevinarus, director and founder of the Inclusive Design Research Centre at the Ontario College of Art and Design University advocates for the virtuous tornado: an inclusive design process that brings in people with disabilities, especially those least served by existing resources, as co-designers guiding the process with lived experience. 

The AI-Powered Future for Learners with Disabilities 

For Jillian, her tech-enabled devices are more than tools—they are her voice to the world and a pathway for exploring her interests and communicating with others. She has used AI to help plan family trips, researching destinations and identifying accessible options. Her dad Mick jokes that she should be a travel advisor for people with disabilities. Jillian’s experience illustrates how technology can open opportunities for self-expression, independence, and a future career. 

To ​ensure that AI truly benefits students like Jillian and Harper, design, implementation, and policy must be research-based and aligned around inclusion, accessibility, and professional development for educators. Prioritizing Students with Disabilities in AI Policy, a paper co-produced by New America and the Educating All Learners Alliance, aims to ensure that new definitions and priorities around AI are intentionally designed for all students. 

Policymakers should set guardrails to make sure that technologies are unbiased and inclusive of diverse perspectives. School districts should prioritize teacher training and opportunities to use AI for innovation, evaluating applications by their ability to support student learning. Developers, meanwhile, must co-design with learners with disabilities and their families, embedding accessibility from the start rather than as an afterthought. Researchers need to continue exploring what works. 

As schools race to adopt AI, the same tools that can supercharge learning also risk amplifying bias, surveillance, and privacy harms. But used with clear guardrails and human oversight, AI can be a lifeline—powering personalized support for learners with disabilities and reducing the educational barriers they face.

Follow The Thread! Subscribe to The Thread monthly newsletter to get the latest in policy, equity, and culture in your inbox.