Enhancing inclusive education in the UAE: Integrating AI for diverse learning needs.
AI chat can boost participation in inclusive classrooms, but only if you block confirmation bias and overload with tight turn limits and live teacher checks.
01Research in Context
What this study did
El Naggar et al. (2024) tested an AI chat system that runs classroom discussions in UAE inclusive classrooms. The tool lets each student ask and answer questions at their own pace. Participants were mixed exceptional learners with diagnoses such as autism, ADHD, and learning disabilities. The team used interviews and classroom notes to see how kids and teachers used the tool every day.
What they found
Kids liked getting personal follow-up questions from the AI. Some spoke more often than in regular whole-class talks. Yet the same feature created problems: a few students kept asking the AI questions until they got answers that matched what they already believed. Others felt lost when the chat stream grew too long. The authors call these downsides confirmation bias and information overload.
How this fits with other research
Brandsen et al. (2024) sounds the alarm first. They show that popular AI language models tag the phrase 'I have autism' as more negative than 'I am a bank robber.' That seems to clash with Alia's hopeful classroom story. The gap makes sense once you see the difference: Sam tested the raw model, while Alia used a teacher-guided chat wrapper that filtered replies.
Herrero-Martín et al. (2024) add a friendly parallel. They used short behavioral profiles to tailor robot-plus-tablet lessons for ASD students. Both studies chase the same goal: tech that adapts to each child in an inclusive room. Herrero-Martín saw higher engagement; Alia saw the same lift, plus new risks.
Fyke et al. (2021) remind us that every tool must tie back to FAPE. After the Endrew F. ruling, IEP goals must be ambitious and evidence-based. If you add AI chat to a goal, you need weekly data showing it helps, not hurts, progress.
Why it matters
You can start using AI chat tomorrow, but treat it like a new paraprofessional. Set clear turn limits so the chat stays short. Watch for kids who hunt until they hear what they want. Log correct and incorrect answers each session. If bias or overload appears, pause the tool and review the transcript with your team. Used with these guards, the same AI that ranks 'autism' as negative can still give your students a louder voice in class.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Open one AI chat thread, set a three-question cap per student, and tally correct versus biased answers for one lesson.
02At a glance
03Original abstract
INTRODUCTION: Artificial Intelligence (AI) mediated systems have become important in educational set-ups, and it is still debatable whether or not they can be useful in special needs education. This qualitative research scrutinizes the experiences and perceptions of exceptional learners, as an example of special needs education, engaged in AI-mediated discussions versus traditional classroom dialogues. The study aims to reveal how these learners process and construct knowledge differently when AI is incorporated into their discussions and how it compares to conventional learning environments. METHODS: The methodology entailed a detailed qualitative analysis, drawing upon cognitive psychology to assess how exceptional learners process information and engage in higher-order thinking. Data were collected through interviews, observation, and content analysis of AI-mediated discussions. RESULTS: FINDINGS: from the study highlighted the capacity of AI technologies to offer personalized and intellectually stimulating educational experiences that resonate with constructivist approaches, promoting active learning and tailored instruction for exceptional learners. However, the research also brought to light certain challenges, including the tendency for confirmation bias and the risk of information overload within AI-mediated environments, which can complicate the learning trajectory within the zone of proximal development. DISCUSSIONS: The study underscores the dynamic interplay between AI technologies and educational processes for exceptional learners. It suggests that while AI can enhance personalized learning, it also introduces unique challenges that must be navigated carefully. Ultimately, this research lays a theoretical and empirical groundwork for the thoughtful integration of AI in supporting inclusive education, emphasizing the importance of continuous evaluation and adaptation.
Research in developmental disabilities, 2024 · doi:10.1016/j.ridd.2024.104685