top of page

A New Institute Is Harnessing AI To Improve Learning For Exceptional Children

For years, there has been a shortage of school-based Speech and Language Pathologists. These professionals provide critical services for students with language disorders, as well as for students with autism, dyslexia, and other disabilities. With school districts struggling to staff their speech and language pathologist positions, many of the 7.3 million students with disabilities in America are at risk of falling behind in their academic and social-emotional development. 


How can schools provide these students with the support they need when they don’t have enough practitioners? This question led to the creation of the National AI Institute for Exceptional Education which is working in partnership with the National Science Foundation and the Institute of Education Science to investigate how AI can be harnessed to support special educators and their students. 

The Institute is supported by a five-year, $20 million grant from NSF and IES awarded in 2023. It is housed at the University at Buffalo and is a partnership of multidisciplinary faculty from eight other institutions: Cornell University, Penn State University, Stanford University, University of Illinois Urbana-Champaign, University of Nevada-Reno, Georgia Institute of Technology, University of Texas at El Paso, and the University of Washington. 

NSF Director Sethuraman Panchanathan recently visited the University at Buffalo to celebrate the Institute’s launch. "Through this new AI Institute, NSF is reinforcing its commitment to our nation's global leadership in AI and in R&D, here in Buffalo and across the country.”

Panchanathan said. “Together, the NSF-led AI Institutes are harnessing the power of AI to strengthen our nation's education system, grow a skilled and diverse workforce, boost the economy and bolster national security." 


The federal funding that supports this AI Institute not only creates opportunities for experts in the field to collaborate, but it also signals to educators, families, speech therapists, and the public that the federal government is invested in finding solutions to a long-standing problem that affects vulnerable students.

National AI Institute for Exceptional Education receives key support from a $20 million grant jointly provided by the National Science Foundation and Institute of Education Sciences.

Building Robot

“This is a national crisis and a topic of importance to so many people,” said Dr. Jinjun Xiong, the Institute’s Scientific Director. “Family members reach out and say that the work being done resonates so much and that they appreciate the federal government funding these innovations.” 


While the Institute is still in its early stages, lead researchers have identified two issues that impede the delivery of school-based speech and language services. 


The first is found within the screening process. The lack of a universal screening system means that some students are identified late. These students require much more support once they are identified than those who received an early intervention. This adds additional stress to a system that is already stretched thin. A similar cycle is seen within intervention: Without enough speech pathologists in schools, it is common for students to receive inadequate individualized and ability-based support or intervention. This means that down the road, students may require even more support, putting even more stress on the already limited system. 


So far, the Institute has developed two umbrella projects to interrupt and address these problems: the AI Screener and AI Orchestrator. The AI Screener will allow for universal early screening for all children, and the AI Orchestrator will be used by teachers and SLPs to provide individualized interventions for students who already have Individualized Education Programs.


“We really hope this funding can make this into a product to put into the hands of teachers, schools, districts, and daycare centers, so people can get the benefit of universal screening.” Dr. Xiong said.


Utilizing these technologies, early identification and intervention will become more accessible without putting further strain on speech pathologists and teachers. The goal is for these systems to alleviate stress on the system while improving the support that is offered to students.


While AI can be useful in providing individualized interventions as a part of a student’s special education plan, it can also be used to develop a more robust system of activities for students needing to strengthen their skills. While observing how speech therapists interacted with students, Dr. Xiong’s research team saw them using flashcards to address a specific speech language disorder, such as a speech sound disorder. Each flashcard contained the word that was difficult for the student. The student worked with the therapist to practice that sound together with other contrasting sounds, under different sound properties. 


After observing these interactions, Dr. Xiong and his team spoke with the therapist who conducted the activity. They learned that it was difficult for them to create the right flashcards for each student’s unique needs, as there was an entire closet of flashcards with different words and sound properties to choose from. The team recognized AI could help solve this problem. Using generative AI, they developed an interface for therapists to input what kind of flashcards they needed. The platform then provides feedback and direction on which words and sounds are most appropriate for that particular student. The prototype has been praised for its ease of use and ability to make sound properties more explainable. 


Depending on how the research progresses, Dr. Xiong is hopeful that some of the innovations from the umbrella projects of the AI Screener and AI Orchestrator will start to roll out for public use in 2026. By the end of 2028, the researchers plan to have proven the benefits of the AI Institute to secure additional funding and keep this crucial work going. 


Without the financial support of NSF and IES, the AI Institute for Exceptional Education would not exist – and the important work that is being done would be impossible. If these AI innovations can be brought to scale, millions of students with disabilities will receive support that they would not have access to otherwise. Federal investments in education research and development will help to fill labor shortages for specialties like speech and language pathology and improve their productivity, and continued investment in this Institute will signal to the public that education, especially for students with special needs, remains one of our nation’s top priorities. 

bottom of page