As a SEND (special education needs and disabilities) coordinator and teacher trainer, I often hear the same questions from colleagues: How do we support learners with dyslexia, ADHD, autism, or speech and language needs in ways that feel realistic in our classrooms? Increasingly, teachers are also asking if generative AI (GenAI) tools have a role to play in helping these learners.
In this article, I'll share some practical ways teachers can use artificial intelligence (AI) and generative artifical intelligence (GenAI) tools (tools that can be used to generate text, images videos or create other media) to support neurodivergent learners, based on real classroom experience. These ideas are rooted in Universal Design for Learning (UDL), a framework that encourages flexibility in how we teach, engage and assess learners.
What is Universal Design for Learning?
Universal Design for Learning, or UDL, is a way of planning lessons that includes all learners from the beginning. Instead of changing a lesson after we see a learner struggling, UDL helps us think ahead so that every learner has what they need to succeed.
UDL is based on three main ideas:
- engagement – How can we keep learners interested and motivated? This might mean giving them choices, using real-world examples or allowing collaboration with others.
- representation – How can we present information in different ways? Some learners understand better by reading, others by listening, watching a video or using images.
- action and expression – How can learners show what they have learned? Not everyone expresses themselves the same way. Some may write an essay; others might give a presentation, make a video or build a model.
As educators, we all want our learners to thrive and feel included. By thinking about these three areas, we create flexible learning environments where all learners can do this. Let's dive into some examples and explore how this might look in real classrooms.
1. Learners with social communication differences
Have you ever had a learner struggle with what to say in a conversation? AI tools like ChatGPT can simulate real-life social scenarios. This gives learners a safe space to practise and build confidence before they encounter these situations in real life.
GenAI tools: Use GenAI tools like Microsoft Copilot or other open-source assistants to simulate conversations, for example:
- managing disagreements respectfully
- role playing social interactions
- scripting responses to common scenarios
Scenario: Amira, a 14-year-old autistic learner, uses an AI chatbot to practise polite ways to ask for help and respond to classmates. She rehearses before class, reducing anxiety. This provides a safe, repeatable environment for rehearsal before real-world interaction and builds on strong organisational skills and the ability to script and prepare ahead.
UDL link: Here GenAI is used to provide multiple means of engagement, that is, diverse options for learners to engage with learning activities. In this case, providing a different way to practise speaking can help learners to manage social stressors proactively.
2. Helping learners stay organised
For learners who find planning and prioritising overwhelming, AI can be a lifesaver. With teacher guidance, GenAI tools can help learners with ADHD or executive function challenges.
Target group: Learners with ADHD or executive function challenges
GenAI can:
- be a space to 'info dump' or use a 'wait and write' technique(Text-based GenAI):
Tool example: Google Keep notes
Some learner, such as those with ADHD or processing difficulties, may interrupt or go off-topic due to challenges with impulse control and attention. These behaviours stem from executive function differences and fall under social, emotional and mental health needs. - create structured outlines or scaffolded prompts for written work or projects (text-based GenAI)
Tool example: ChatGPT/Custom GPTs
Helps reduce cognitive load and supports planning by breaking tasks into manageable chunks, making it easier for learners with ADHD to focus and get started. Prompts can be personalised or simplified for clarity. - create visual checklists:
Tool example: Canva AI or Google Keep
Provides clear, step-by-step visual guidance that supports working memory and helps learners stay on track. Customisable templates with icons or colour coding reduce overwhelm and promote independence. - prompt timers and reminders:
Tool example: AI-powered apps like Google Assistant, Siri or Time Timer app.
Aids time management and attention regulation by offering gentle cues to start, continue or wrap up tasks. These tools can provide voice or visual prompts and support routines and transitions.
Scenario: Liam, a 17-year-old with ADHD, uses a planning AI to structure his week. The AI breaks down his homework into smaller steps, helping him manage deadlines without overwhelm.
UDL link: Here AI provIdes multiple means of action and expression, giving learners flexible tools to organise and complete tasks.
3. Making written content more accessible
Reading can be tough for some learners, but AI-powered tools can make text more accessible.
Target group: Learners with dyslexia, processing difficulties or visual impairments
Use AI to:
- summarise a grammar focus in easier-to-understand language
- restructure and give feedback on writing
- work alongside text-to-speech tools, so that learners can listen instead of reading; for learners with processing difficulties, this removes a huge barrier to learning
- allow learners to contribute ideas using multiple formats: text, voice recordings, images or videos using platforms such as Padlet.
Scenario: Raya, a 14-year-old, uses an AI assistant to summarise her reading texts and dictate written responses via voice input. This allows her to focus on ideas rather than spelling. Makes reading, writing and understanding content more accessible.
UDL link: By combining text, image and sound, AI can provide multiple means of representation.
4. Speech and communication needs
Learners with autistic spectrum conditions (ASC) or stammering may experience challenges with expressive or receptive language, pragmatic (social) communication or fluent speech. These difficulties can impact classroom participation, peer interaction and confidence, especially in spoken tasks.
Autistic learners, for example, may engage in what's known as 'info-dumping', speak with flat intonation or struggle with turn-taking and figurative language. Learners who stammer may avoid speaking entirely for fear of judgement or interruption.
Early support from speech and language specialists, paired with inclusive classroom strategies, can greatly enhance communication and engagement.
Target group:
- Learners with ASC, stammering or other speech/pragmatic language differences
- Learners who demonstrate reluctance to speak, difficulty initiating conversation or anxiety in communication-heavy tasks
GenAI support:
- Speech-to-text: Reduces anxiety during oral presentations or group work (e.g. Otter.ai, Microsoft Dictate)
- Text-to-speech: Supports receptive language or self-editing written work (e.g. NaturalReader, Immersive Reader)
- Rewording and clarification tools: Help students express themselves more clearly in writing or speech (e.g. ChatGPT)
- AI-generated sentence starters: Scaffold communication and written expression for those with expressive difficulties
Mini case study – Learner with ASC
Noah, a 14-year-old learner with autism, finds it difficult to express thoughts clearly in class discussions. He often info dumps when excited and struggles to summarise ideas when writing. His teacher introduces ChatGPT to help him summarise longer responses and generate sentence starters for writing tasks. Noah also uses speech-to-text tools when drafting ideas, allowing him to speak naturally and edit afterwards. Over time, his written and spoken contributions become more focused, and he engages more confidently with group tasks.
UDL link: Giving learners flexible tools and formats (speech, text, visual) is another way of offering multiple means of expression, and allows them to express what they know in a way that suits their communication profile.
5. Processing speed and memory
Slow processing speed and memory difficulties can arise from various factors, including brain differences, attention challenges, learning disabilities, and emotional or physical conditions. Effective interventions involve utilising strategies to improve processing speed and memory skills.
Target group: Learners who struggles to retain, retrieve or process new information quickly
GenAI support:
- Quizzes for recall practice (Curipod or Kahoot! AI for recall-based quiz games)
- Summarise lengthy materials
- Visual AI tools for memory
Scenario: Ahmed uses a quiz generator to create flashcards for vocabulary review. He reviews them in short intervals and builds confidence.
UDL Link: Using AI tools for gamified repetition helps supports learners' memory.
But what about the risks of using AI tools?
AI isn't perfect. We need to be mindful of ethical concerns, including:
- bias in AI models, which could reinforce stereotypes
- privacy concerns, especially when handling student data
- over-reliance on AI, which should support, but never replace, the expertise of teachers and SEND professionals
- the importance of a human-first approach that prioritises learner needs, including better learning outcomes; if the use of AI isn't enhancing learning, it isn't the best approach.
Final thoughts: Balancing AI with human expertise
Neurodivergent learners face barriers not because of their abilities but because systems often aren't designed for them. AI tools offer opportunities to personalise learning, reduce stress and build independence, when used mindfully and ethically.
AI won't replace teaching, but it can amplify inclusive practice, giving all learners a chance to succeed on their terms.
Download the template below to support you when you start experimenting with AI tools. It encourages you to reflect without making it overwhelming. This AI Pilot Log has been designed in alignment with the graduated approach to SEND support, as outlined in the SEND code of practice (0–25). Each section of the log mirrors the four-part cycle of Assess, Plan, Do, Review, supporting teachers to reflect on how AI tools are being used to meet the diverse needs of learners.
Download the AI Pilot Log
Nadia Cooke is a Special Education Needs and Disability Coordinator and teacher trainer. She has a particular interest in AI and SEND and has delivered workshops and training sessions at local and international level. Her work focuses on supporting teachers in developing practical classroom strategies that enhance learning and communication.