Transforming Special Education: The Role of Technology, Especially

AI, in Enhancing Inclusivity and Learning Outcomes

Meirun Wang

Institution of Education, University College London, London, WC1E 6BT, U.K.

Keywords: Special Education, AI in Education, Personalized/Adaptive Learning, ASD Intervention, Equity & Inclusion.

Abstract: This paper explores the transformative role of Information and Communication Technologies (ICTs) and

Artificial Intelligence (AI) in special education, focusing particularly on their potential to foster inclusivity

and equity in education for disabled individuals. Globally, over one billion people experience some form of

disability, often facing social exclusion and discrimination, which can be mitigated through targeted

educational strategies. The integration of advanced technologies like AI and VR in educational systems offers

unprecedented, personalized learning experiences, addresses the diverse needs of disabled students, and

supports their integration into society. This study delves into the applications of these technologies globally

and in China, highlighting the challenges and potential solutions to ensuring quality education for all. Through

the analysis of current technological applications and their impact, the paper discusses how digital tools and

AI-driven solutions not only enhance learning outcomes but also promote social and cognitive inclusion.

1 INTRODUCTION

According to data from the UNESCO Asia Pacific

Education Bureau and the World Health Organization

(WHO), more than one billion people worldwide

experience some form of disability, and this figure

continues to rise as the global population increases

(World Health Organization, 2018). Due to the

unique behavioral and cognitive characteristics of

individuals with disabilities, they often face various

degrees of discrimination and marginalization in

society. This societal exclusion not only threatens

social stability but also infringes upon the equal rights

of citizens (Ditchman et al., 2016). Therefore, using

educational measures to support and advocate for this

group is increasingly important for both national and

societal well-being. Developing specialized

education for disabled children is a crucial strategy

for promoting equal educational opportunities (Miles,

2012). Assisting disabled children in integrating into

their peer groups through special education and

fostering their holistic development is a societal issue

that warrants attention.

China has the largest population of disabled

individuals in the world (Guo, 2014). The Chinese

government has highlighted a focus on ensuring the

enrollment rate of individuals with disabilities in its

development plans for special education. However,

amidst high enrollment rates in inclusive education

settings, ensuring high-quality, effective education

and adequate educational resources remains a

challenge (Deng & Harris, 2008). Additionally, given

the unique needs and educational goals of each

disabled child, particularly in the context of a

shortage of qualified special education teachers and

limited professional training, meeting current special

education needs is difficult. Thus, the development of

information technology in special education and the

expansion of resources are critically important. These

advances not only help integrate individuals into

social life but also provide personalized learning

experiences that support individualized learning. This

article will detail the potential and applications of

technology and AI in the field of special education.

2 THE IMPACT OF

INFORMATION

TECHNOLOGY ON SPECIAL

EDUCATION

The rapid development of ICTs has significantly

altered the living conditions of many individuals. The

application of ICTs in the field of education has also

amassed substantial evidence to date. These

Wang, M.

Transforming Special Education: The Role of Technology, Especially AI, in Enhancing Inclusivity and Learning Outcomes.

DOI: 10.5220/0014003900004912

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Innovative Education and Social Development (IESD 2025), pages 593-598

ISBN: 978-989-758-779-5

593

technologies provide relatively personalized and

timely feedback. The introduction of multimedia

technologies has greatly enriched teaching methods,

enhancing students' learning motivation and

classroom engagement. For students with autism and

attention deficits, modern technological devices like

tablets with touchscreen capabilities significantly aid

their understanding and interaction.

Dating back to the early 1980s, ICT specifically

designed for students with learning disabilities had

already emerged. The application of this technology

heavily depended on the exceptionality it was

designed for, leading to considerable variability. For

students requiring special education, the predominant

technology used comprised computer-assisted

instruction (CAI) (Jeffs et al., 2003). The primary

purpose of this instruction was to develop specific

skills through repetitive practice, viewing the

computer as a teacher's role and providing timely

feedback to students. Nevertheless, Hummel et al.

(1985) critiqued that CAI also had its limitations,

lacking significant interaction and serving merely as

an aid. This aspect was particularly unfriendly to the

learning process of students with learning disabilities,

as their attention could easily be diverted. Without

supportive, interactive content, there tends to be a

gradual reduction in students' motivation and

engagement, fostering a traditional 'computer-

centred' passive learning environment.

As technology progressed in the late 1980s and

early 1990s, researchers in the field of learning

disabilities started investigating the use of graphics

and multimedia in education. Multimedia, which

combines elements like graphics, videos, animations,

images, and sound, provides a variety of instructional

approaches. The Cognition and Technology Group at

Vanderbilt (CTGV) carried out extensive research on

multimedia teaching methods, with a particular

emphasis on videodisc environments. These

contextualized settings provided learners with

valuable opportunities to actively build knowledge

within a realistic learning context (Cognition and

Technology Group at Vanderbilt, 1993). It is regarded

as a method that allows students to connect their

unique perspectives and expressive methods with the

common curriculum (Najjar, 1996). Increasingly,

multimedia applications have transitioned to

computer-based platforms, shifting from passive

reception to more interactive modes. The research

conducted by Daiute and Morse (1994) involved the

use of multimedia writing tools to assist low-

achieving and reluctant writers, of whom five-sixths

required special education services. The study found

that multimedia learning materials attract learners

through various forms of presentation, thereby

supporting the positive benefits of multimedia

instruction for students with learning disabilities or

those lacking prior knowledge in specific academic

areas.

Furthermore, an educational support centre in

Western Australia has also enhanced practical and

perceptual learning by incorporating iPads into

teaching. By equipping each classroom with two

iPads and frequently using the devices in teaching,

personalized educational plans incorporating iPad

applications were developed based on teachers'

understanding of individual student needs. According

to classroom teachers' feedback, the large

touchscreens and swipe functions of the iPads are

particularly beneficial for children with motor control

issues (Johnson, 2013). The educational trials have

shown promising results in enhancing student

motivation, especially among students with autism

and attention deficit disorders (Johnson, 2013).

3 CHALLENGES AND

LIMITATIONS: THE

APPLICATION OF

TECHNOLOGY IN

EDUCATION AND ITS SOCIO-

ECONOMIC IMPACTS

However, while this instructional method has

enhanced teaching outcomes, it also faces challenges

such as insufficient device availability and potential

social isolation (Johnson, 2013). Interactions with

machines do not improve students' social skills in real

life; instead, prolonged use may lead to increased

seclusion, exclusion, and isolation in interpersonal

interactions. Additionally, although these devices

provide a certain level of personalization, the

teaching process still largely depends on the teacher's

knowledge of each student and long-term surveys of

volunteers. This could lead to problems of self-

presentation bias and delays in information

availability (Kopcha & Sullivan, 2006).

Additionally, the financial burden required to

support these devices is not something every school

can afford. The reliance on technological resources

may also exacerbate educational inequalities

(Rafalow & Puckett, 2022), particularly in

economically disadvantaged areas where students

may be unable to access necessary technological

devices, thus missing out on the benefits of

multimedia education. These issues indicate that

IESD 2025 - International Conference on Innovative Education and Social Development

594

technology must undergo further innovation to

address the existing challenges in education.

4 THE ROLE OF AI

TECHNOLOGY IN THE

DIAGNOSIS AND EDUCATION

OF CHILDREN WITH SPECIAL

NEEDS

The integration of AI with education offers a new

perspective on these issues and has now become a

core technological domain supporting formal

education and lifelong learning (Luckin & Holmes,

2016). AI in education represents a new field that

emerges from the intersection of artificial intelligence

and educational theory. It aims to merge AI with

education, employing advanced technological means

to enhance teaching quality, ensure greater equity in

education, and diversify educational support (Han et

al., 2022). Currently, AI is predominantly applied in

the diagnostic and screening processes to evaluate

children with autism and learning disabilities and to

send reports to doctors and parents. The objectivity

and accuracy of these assessments are enhanced by

their deep-learning capabilities and machine-learning

algorithms (Lu & Perkowski, 2021). These

applications demonstrate that AI now has the capacity

to distinguish between children who require special

education and those who are suitable for mainstream

education. This ability signifies that AI has mastered

the characteristics of the special education

demographic. Identifying these needs lays the

foundation for providing appropriate support in

aspects of life, social interaction, and personalized

teaching strategies for this group of students.

5 GLOBALLY & CHINA'S AI AND

SPECIAL EDUCATION

RESEARCH

5.1 Support in Social Life

In the area of enhancing social skills and intervening

in the social behaviour disorders of individuals with

autism spectrum disorder (ASD), numerous studies

have found that children with ASD often show a

greater interest in robots or other forms of AI. Simut

et al. (2015) found that children with autism often

made more eye contact with AI robots and preferred

interacting with AI rather than caregivers and peers.

Shi (2019) suggests that this phenomenon is due to AI

being less intrusive and simpler compared to human

interactions. These characteristics make AI an ideal

tool for improving the social skills of children with

autism. Numerous studies indicate that interactions

between AI robots and children with autism can

enhance their joint attention skills, which are

considered one of the most crucial aspects of social

abilities. Additionally, AI can enhance ASD

children's ability to be oriented to prompts and

attention (Warren et al., 2013).

In a study conducted in China, AI was utilized as

a reinforcement tool within Applied Behaviour

Analysis (ABA) interventions. The research focused

on the intervention combining a humanoid robot

named Wukong with ABA training for a 7-year-old

child with autism. The results demonstrated that

through interaction with this AI robot, Wukong

quickly became the most effective reinforcer during

the study, not only enhancing the creativity and

interaction of the autistic children but also serving as

a bridge for communication between them and their

neurotypical peers. For instance, the study noted that

the familiarity of autistic children with Wukong's

functions earned the admiration of their peers for their

ability to control the robot. This represents the first

step in integrating ASD children into the mainstream

educational system. It acts as a link between them and

typically developing children, thereby providing

more opportunities and possibilities within an

inclusive educational environment (Shi, 2019).

Despite these benefits, the research also

highlighted some limitations and potential risks

associated with Wukong. For example, due to the

limited interactive modes of the AI robot, it struggles

to capture the emotional fluctuations of children with

autism. Moreover, a two-month trial study revealed

that children with autism might develop a dependency

on the AI robot, consequently neglecting

communication with their parents. This dependency,

coupled with a fixed pattern of communication, may

lead to self-stimulating behaviors such as echolalia

and avoiding interactions with others. However, Shi

(2019) points out that compared to evidence-based

interventions, AI is seen as an effective tool for

reducing high costs, particularly in enhancing the

social skills of children with autism. Given that AI

robots like Wukong currently cannot autonomously

adjust and mitigate these negative impacts, the

research suggests that AI robots can only serve as an

effective reinforcer under manual control. Future

Transforming Special Education: The Role of Technology, Especially AI, in Enhancing Inclusivity and Learning Outcomes

595

research and practice will need to focus on how to

harness the potential of AI for children with autism

while avoiding negative impacts.

5.2 Support in Personalized Learning

Children with special needs exhibit significant

learning differences based on the type and severity of

their disabilities. However, in the present one-to-

many educational environment, personalized tutoring

is often limited, causing teachers to neglect the

specific learning support needs of individual students,

which leads to an unequal allocation of educational

resources (Deng & Pei, 2009). The integration of

artificial intelligence and neural networks provides an

unprecedented, personalized learning solution for

education. AI-based intelligent teaching systems

employ algorithms and data, integrating

computational intelligence, learning analytics, and

data mining techniques (Han et al., 2022). These

systems can track students' learning dynamics in real-

time, collecting and analyzing past learning states and

progress. Based on this information, the systems

comprehensively understand each student's learning

level and needs, tailoring adaptive learning plans

accordingly. Moreover, throughout the teaching

process, the system can automatically generate

appropriate questions and answers based on the

learning goals and individual student levels, adjusting

the content and pace dynamically by assessing the

accuracy of student responses. This approach aims to

enhance educational efficiency and ultimately

supports student learning and assists teachers in their

instructional roles (Han et al., 2022).

For students with visual impairments who are

unable to acquire knowledge through sight, learning

must rely on tactile or auditory methods. The

incorporation of artificial intelligence and neural

networks supports the collection, processing, and

analysis of individual learning data, allowing for the

creation of customized instructional programs (Tan &

Wang, 2020). This approach also involves the use of

brain-computer interface (BCI) technology to

reconstruct neural signals in the visual cortex,

assisting in their learning process. Additionally, for

students with expressive disorders, developed BCI

technology can directly collect and analyze brain

signals to assess whether the students are focused or

if their emotional states are stable (Jiang et al., 2018).

Teachers can use this real-time information to adjust

teaching strategies and provide targeted instruction.

Students with specific learning disabilities or

language disorders require assistance in reading,

writing, pronunciation, and comprehension

(Sarisahin, 2020). The BeeSpecial software platform

uses artificial intelligence to deliver personalized

digital tools that help students with dyslexia, reducing

the challenges they usually face and supporting their

academic success (Zingoni et al., 2021). The

platform's implementation occurs in stages. Initially,

clinical reports on dyslexia, self-assessment

questionnaire responses, and results from a series of

psychometric tests are entered into the system. AI

processes this data to extract crucial information

regarding student needs and challenges, forming an

initial predictive model. This model forecasts the

most suitable support methods for each student,

outlining best practices for teachers and educational

institutions and ways to make learning materials more

accessible through digital tools. In the second phase,

each student tests the digital support tools, and their

responses, improvements, and remaining challenges

are evaluated. These assessment results are fed back

to the AI, transforming the predictive model from

category-specific to student-specific.

Additionally, virtual reality (VR) technology will

be applied in the assessment module, as it can present

materials in a more engaging manner and easily

collect the necessary information. This approach not

only helps mitigate issues of dyslexia and attention

deficits but also simulates the challenges faced by

students with reading difficulties, enabling teachers to

understand these phenomena better and intervene

appropriately (Zingoni et al., 2021). The platform has

already collected data from approximately 700

students with dyslexia and carried out preliminary

analyses, yielding initial results about the most

significant challenges and the most effective support

tools and strategies.

Attention problems and memory impairments are

the main difficulties faced by these students (Zingoni

et al., 2021). To address these issues, the platform

emphasizes the need for proper support and has

identified that using highlighted keywords, clear

layouts, along with images, summaries, concept

maps, and diagrams are the most effective tools.

Furthermore, strategies such as pausing during

classes, hosting online sessions, repeating learning

materials, and providing course programs and

slideshows are considered most appropriate,

significantly enhancing learning efficiency. The

platform also advocates utilizing auditory channels,

such as recording lessons, using audiobooks, and

preferring oral exams, to cater to specific student

needs. These results will serve as valuable guides for

refining the concept and technological choices of the

BeeSpecial platform. The next steps in

implementation will involve training the AI module

IESD 2025 - International Conference on Innovative Education and Social Development

596

to automatically predict the most useful tools and

strategies, as well as incorporating VR capabilities to

administer psychometric tests and assess digital tools.

6 CONCLUSIONS

The deployment of Technology and AI in special

education presents a vital opportunity to revolutionize

how educational services are delivered to disabled

students. As demonstrated in the various case studies

and literature reviewed, these technologies provide

critical support in personalizing learning experiences,

enhancing student engagement, and facilitating the

social integration of students with disabilities.

However, the effective implementation of such

technologies requires overcoming significant

challenges, including the equitable distribution of

educational resources, professional training for

educators, and the development of infrastructure to

support technology-driven teaching methods.

Moreover, future research should focus on refining

AI models to better address the nuanced needs of

disabled students and expanding the use of VR to

simulate complex learning environments. By

advancing these technologies, educators can

significantly improve the educational landscape for

students with disabilities, making it more inclusive

and effective. The continued evolution of AI and VR

in special education holds the promise of creating

more equitable educational opportunities and

fostering a more inclusive society.

REFERENCES

Cognition and Technology Group at Vanderbilt. (1993).

Examining the cognitive challenges and pedagogical

opportunities of integrated media systems: Toward a

research agenda. Journal of Special Education

Technology, 12(2): 118–124. https://doi.org/10.117

7/016264349301200204

Daiute, C., & Morse, F. (1994). Access to knowledge and

expression: Multimedia writing tools for students with

diverse needs and strengths. Journal of Special

Education Technology, 12(3): 221–256.

https://doi.org/10.1177/016264349401200305

Deng, M., & Harris, K. (2008). Meeting the needs of

students with disabilities in general education

classrooms in China. Teacher Education and Special

Education: The Journal of the Teacher Education

Division of the Council for Exceptional Children,

31(3): 195–207.

https://doi.org/10.1177/0888406408330631

Deng, M., & Pei, M. (2009). Instructions for students with

special educational needs in Chinese mainstream

classrooms: Modifications and barriers. Asia Pacific

Education Review, 10(3): 317–325. https://doi.org/10.

1007/s12564-009-9032-1

Ditchman, N., Kosyluk, K., Lee, E. J., & Jones, N. (2016).

How stigma affects the lives of people with intellectual

disabilities: An overview. In K. Scior & S. Werner

(Eds.), Intellectual disability and stigma (pp. 31–47).

Palgrave Macmillan. https://doi.org/10.1057/978-1-

137-52499-7_3

World Health Organization. (2018). International

classification of diseases for mortality and morbidity

statistics (11th ed.). World Health Organization.

Guo, C. (2014). Development of disability policy in China.

China Journal of Social Work, 7(2): 208–210.

https://doi.org/10.1080/17525098.2014.921763

Han, X., Hu, L., Han, D., Peng, Y., Wang, Y., Yan, C., &

Wang, Z. (2022). Research on the application of

artificial intelligence in special education. In

International Conference on Social Science, Education

and Management.

Hummel, J. W., Farr, S. D., Hazan, P., Zuckerman, R., &

Lamos, J. P. (1985). Options for creating and modifying

CAI software for the handicapped. Journal of Learning

Disabilities, 18(3): 166–168. https://doi.org/10.1177/0

02221948501800313

Jeffs, T., Morrison, W. F., Messenheimer, T., Rizza, M. G.,

& Banister, S. (2003). A retrospective analysis of

technological advancements in special education.

Computers in the Schools, 20(1–2): 129–152.

https://doi.org/10.1300/j025v20n01_10

Jiang, L., Zhang, H., Zhang, L., Wu, B., Sun, Q. C., & Li,

H. B. (2018). Knowledge mapping analysis of the

evolution of brain-computer interface research and

trends in educational applications: A study based on

SCI and SSCI journal papers from 1985–2018. Journal

of Distance Education, 36(4): 27–38.

Johnson, G. (2013). Using tablet computers with

elementary school students with special needs: The

practices and perceptions of special education teachers

and teacher assistants. Canadian Journal of Learning and

Technology, 39(4). https://www.learntechlib.org/p/

130195/

Kopcha, T. J., & Sullivan, H. (2006). Self-presentation bias

in surveys of teachers’ educational technology practices.

Educational Technology Research and Development,

55(6): 627–646. https://doi.org/10.1007/s11423-006-

9011-8

Lu, A., & Perkowski, M. (2021). Deep learning approach for

screening autism spectrum disorder in children with

facial images and analysis of ethnoracial factors in model

development and application. Brain Sciences, 11(11):

1446. https://doi.org/10.3390/brainsci11111446

Luckin, R., & Holmes, W. (2016). Intelligence unleashed: An

argument for AI in education. Discovery.ucl.ac.uk.

https://discovery.ucl.ac.uk/id/eprint/1475756/

Miles, S. (2012). Justice and equality in education: A

capability perspective on disability and special

educational needs. International Journal of Disability,

Transforming Special Education: The Role of Technology, Especially AI, in Enhancing Inclusivity and Learning Outcomes

597

Development and Education, 59(2): 221–222.

https://doi.org/10.1080/1034912x.2012.676445

Najjar, L. J. (1996). Multimedia information and learning.

Journal of Educational Multimedia and Hypermedia,

5(2): 129–150. https://eric.ed.gov/?id=EJ534457

Rafalow, M. H., & Puckett, C. (2022). Sorting machines:

Digital technology and categorical inequality in

education. Educational Researcher, 51(7): 508–518.

https://doi.org/10.3102/0013189x211070812

Sarisahin, S. (2020). Reading comprehension strategies for

students with learning disabilities who are emergent

bilingual. Intervention in School and Clinic, 56(1): 5–

13. https://doi.org/10.1177/1053451220910731

Shi, J. (2019). The application of AI as reinforcement in the

intervention for children with autism spectrum disorders

(ASD). Journal of Educational and Developmental

Psychology,9(2):17.https://doi.org/10.5539/jedp.v9n2p1

Simut, R. E., Vanderfaeillie, J., Peca, A., Van de Perre, G., &

Vanderborght, B. (2015). Children with autism spectrum

disorders make a fruit salad with Probo, the social robot:

An interaction study. Journal of Autism and

Developmental Disorders, 46(1): 113–126.

https://doi.org/ 10.1007/s10803-015-2556-9

Tan, Y., & Wang, C. (2020). Research on the application of

artificial intelligence in individualized education for

special children. Journal of Suihua College, 40(7): 126–

129.

Warren, Z. E., Zheng, Z., Swanson, A. R., Bekele, E., Zhang,

L., Crittendon, J. A., Weitlauf, A. F., & Sarkar, N. (2013).

Can robotic interaction improve joint attention skills?

Journal of Autism and Developmental Disorders, 45(11):

3726–3734. https://doi.org/10.1007/s10803-013-1918-4

Zingoni, A., Taborri, J., Panetti, V., Bonechi, S., Aparicio-

Martínez, P., Pinzi, S., & Calabrò, G. (2021).

Investigating issues and needs of dyslexic students at

university: Proof of concept of an artificial intelligence

and virtual reality-based supporting platform and

preliminary results. Applied Sciences, 11(10): 4624.

https://doi.org/10.3390/app11104624

IESD 2025 - International Conference on Innovative Education and Social Development

598