Artificial intelligence is increasingly being promoted as a way to make classrooms more accessible to learners whose brains process information differently. Globally, as many as 20% of people are estimated to be neurodivergent — a group that includes learners with dyslexia, autism, attention-deficit hyperactivity disorder (ADHD), dyspraxia and speech or language processing differences — roughly 1.6 billion people.
How AI can change classroom practice
Proponents say AI-powered systems can personalise instruction in real time, providing adaptations that would otherwise demand intensive teacher time and specialist resources. Practical examples include:
- automatically simplifying or restructuring text into shorter sentences and clearer chunks for learners with dyslexia;
- transforming spoken lessons into structured notes or visual summaries for those who struggle with working memory or language processing;
- breaking homework into timed microtasks and reminders for learners with ADHD;
- generating visuals or diagrams that scaffold complex concepts without requiring teachers to prepare multiple differentiated resources.
Advocates emphasise that AI is a tool for teachers, not a replacement. By reducing the time teachers spend manually differentiating materials, these tools can free educators to focus on relationships, classroom management and higher-order instruction.
Cognitive load and accessibility
One clear area of benefit is what specialists call cognitive load reduction. For many neurodivergent learners, decoding dense text, filtering auditory distractions or tracking multi-step instructions consumes considerable mental effort. AI systems that reformat text, add visual scaffolds or convert speech into organised notes can lower that burden and allow learners to concentrate on meaning and problem-solving.
Those benefits can be delivered within a single mixed-ability classroom: AI can present multiple formats simultaneously to match individual preferences without requiring the teacher to produce separate lesson plans for each learner.
Opportunities and questions for South Africa
South African schools face both opportunity and challenge in adopting such tools. On the positive side, AI offers potential to extend scarce specialist supports into under-resourced contexts and to help teachers cope with large class sizes. But three implementation issues are especially pressing:
- Access and equity: Many AI applications require reliable internet, devices and electricity. Without targeted investment, schools in poorer communities risk being left further behind.
- Teacher training: Successful rollout depends on teachers being able to interpret AI outputs, integrate them into lessons and know when to override automated suggestions.
- Privacy and data governance: AI systems often rely on learner data to adapt instruction. Clear policies are needed to protect learners’ personal information and to ensure transparency about how data are used.
Policymakers, district officials and school leaders will need to weigh these factors carefully. Scaling helpful tools without addressing infrastructure and capacity will limit impact and risk widening existing disparities.
Practical steps for schools
Education systems considering AI-supported inclusion could take several pragmatic steps:
- pilot tools in a range of settings to gather evidence on classroom impact and technical requirements;
- invest in teacher professional development that focuses on interpreting AI recommendations and designing inclusive lesson plans;
- develop clear data-protection rules and consent processes for learners and caregivers;
- prioritise offline and low-bandwidth solutions where connectivity is unreliable.
When introduced thoughtfully, AI has the potential to broaden individualised support to many more learners than current specialist services can reach. But realising that promise will require deliberate policy, adequate funding and sustained teacher support so that tools enhance — rather than complicate — day-to-day teaching and learning.
