Artificial intelligence is revolutionising students’ experience of higher education, perhaps particularly in breaking down barriers to learning. As an academic and teacher, I see AI as a tool that can promote inclusion and accessibility, ensuring that all students have more equitable opportunities to engage with and benefit from their studies.

One of the most compelling cases for AI in education is its ability to reduce workload for both staff and students. Generative AI tools can offer drafts, refine writing, and summarise key ideas, saving time. However, beyond workflow improvements, AI offers profound potential in making learning more accessible for students with diverse needs.

Many students face significant barriers with traditional teaching methods. Dyslexic students, for instance, struggle with long and complex journal articles, while non-native speakers may find academic English daunting. AI tools like text-to-speech applications and large language model summarisation can remove these obstacles. A long research paper can be distilled into key points, while screen readers can read aloud any text, helping students focus on comprehension rather than decoding graphemes.

The traditional model of higher education—lectures, seminars, library research, assignments—can be exclusionary. Large language models can function as personal tutors, providing instant, tailored explanations and critiques of students’ work. This mirrors the supervision model at elite universities, where students engage in intensive one-on-one discussions with tutors. AI brings this level of individual support to all learners, not just those at some other collegiate universities.

There are ethical concerns, of course. We must ensure that AI is used to support learning, not replace it. Students should develop their own critical thinking and communication skills rather than rely on AI-generated work. AI can refine a draft or help a student overcome writer’s block, but it should not become a shortcut that bypasses the learning process altogether.

The integration of AI-driven accessibility tools within our much loved Office 365 is another game-changer. Features such as real-time captioning, translation tools, and dictation software help students engage with content in the way that best suits their needs. A student with ADHD who struggles to maintain focus on long readings can use AI to generate concise summaries, making engagement with course materials more manageable.

Moreover, AI-generated audio summaries and podcasts offer an alternative to reading. Some students retain information better when listening rather than reading, and AI makes this adaptation seamless. Tools like Google’s Notebook LM allow students to upload academic papers and generate study guides or engaging podcast-like discussions, helping them grasp complex concepts.

Despite the enormous benefits, AI in education comes with a significant environmental cost. Training and operating large AI models require large computational resources. Universities must balance AI adoption with sustainability considerations, ensuring that these technologies serve students without undue ecological impact.

Ultimately, AI should be a means to empower learners, not a crutch that diminishes their independence. The goal should be to enhance accessibility, promote engagement, and facilitate deeper learning. By thoughtfully integrating AI, higher education can become more inclusive, ensuring that all students, regardless of background or ability, have the support they need to succeed.

*Based on a talk I gave at Roehampton’s Academic Summit, 10 January 2025. *