Eight ways to use AI in active learning – and four challenges it brings
Use AI in active learning by deploying these strategies to integrate tools in your teaching. Plus, how to address four challenges when doing so
How can we leverage AI tools to enhance active learning? AI can enable students to engage more deeply and can support their problem-solving processes. It can also automate tasks, personalise learning experiences and enable real-time feedback.
Active Blended Inclusive Learning means using digital technologies, such as AI, to help learners to engage inclusively with a series of activities, which require them to produce evidence of their learning. Active learning has been shown to have a range of benefits, promoting equity and enhancing student outcomes. Using active learning in our classrooms has led to increased engagement and academic achievement, development of critical thinking and higher order thinking skills, and deeper retention of knowledge.
How to apply AI to promote active learning
Generation of images: Applications like DALL·E, Leonardo.ai and Flux can be used to generate images from simple text, to support discussions and encourage creativity. They also can be used to generate images that may not yet exist. For example, a reconstruction of a marketplace in Ancient Mesopotamia can help bring history to life.
Generation of videos: Many students use YouTube or other videos to support learning. Tools such as Synthesia enable users to create videos incorporating
AI avatars that can present text: Students can convert textual course content to videos, making it more engaging. Videos are simple to edit and can be translated into any language. Open AI’s Sora model can create photorealistic video models and special effects.
Intellectual play and philosophical inquiry: Tools like OpenAI’s voice mode, ElevenLabs’ conversational agents and Google NotebookLM can be used for thought experiments, philosophical dialogues, roleplays, podcasts or scenario-based learning. For example, students can roleplay with a thinker (eg, bel hooks on education or Ada Lovelace on mathematics) or explore a scenario or thought experiment relevant to their field (eg, a play-oriented university or Laplace’s Demon and quantum mechanics).
Generate code: Anthropic Claude, OpenAI ChatGPT and Google Gemini can generate code snippets, suggest improvements and debug errors. Codecademy provides step-by-step coding tutorials, including how to use AI in coding.
Scaffold complex topics: AI tools can generate supporting material to scaffold comprehension of difficult topics, such as mathematical problems, timelines of historical events and explanations of policy documents.
Feedback on writing: Students can use AI tools, such as Chat GPT’s canvas feature and Microsoft Copilot’s integration with Microsoft Word, to get feedback on any specific aspects of their writing (such as clarity, spelling, critical thinking, subject knowledge) without compromising academic integrity. For example, a student could request the following feedback: “Please review my draft for clarity, conciseness, grammar and style. Offer guidance and examples of how I might improve my writing skills but do not rewrite any content.”
Develop critical thinking: AI can help support critical thinking by advising on alternative sources and providing evaluations of these sources. This is more powerful than a web search because it can infer what the user is looking for and suggest optimised keywords for searches. Yes, AI can hallucinate sources, but this provides more opportunities for students to critically evaluate output. With fine-tuning and refined prompts, hallucinations can be reduced to a minimum. In particular, models such as Perplexity and GPT-4o are getting better at avoiding such hallucinations, with the ability to directly link to contemporary source materials and referencing real-world links outside the training data.
Interactive and learning games: Teaching technical terminology can be easily reframed within a responsive fill-in-the-blanks game, rivalling Wordle. Using Anthropic’s Claude for coding, Black Forest Labs’ Flux for visuals, and Google Labs’ MusicFX for sound, educators can create stunning multimedia games that make learning and revision engaging, rewarding and something students actively return to.
- Let’s look at AI as a reasoning partner, not a shortcut
- Tweaks to make when teaching political science and social policies in the GenAI age
- An AI toolkit for all aspects of academic life
Challenges and possible solutions:
- Culture of resistance: There is often resistance from colleagues who may fear new technology. Addressing this requires accessible tools, clear ethical guidelines and training to support staff.
- Data privacy: We need to make ourselves aware of whether tools use our data to train their models, which types of data are appropriate to input and if our organisations have non-disclosure agreements with these technology vendors.
- Accuracy of information: AI tools cannot always provide accurate and impartial output. For now, we always need to have a human in the loop to apply their expertise to confirm the accuracy of AI-generated content.
- Student misconduct issues: Students may simply get AI to do the things their instructors have asked them to do. In the past, in-person exams, penalties and detection tools helped mitigate misconduct, but AI is forcing universities to rethink authentic assessment design.
We’ve suggested a number of ways that AI tools can be used to promote active learning in higher education teaching. We encourage you to be cautious and judicious, but also adventurous in experimenting, innovating and collaborating with your learners to make educational experiences more engaging and inclusive.
Tab Betts is a lecturer in higher education pedagogy at the University of Sussex. Shelini Surendran is an associate professor in biosciences; Martin Hawes is a senior lecturer in veterinary pharmacology and therapeutics; Stella Kazamia is senior lecturer in computer science and electronic engineering; Joey Sikchun Lam is a lecturer in computer science and electronic engineering, all at the University of Surrey.
Shaun Le Boutillier is associate professor of sociology; Uwe Matthias Richter is associate professor and academic lead: digital pedagogic innovation; Jason Williams is media specialist, all at Anglia Ruskin University.
Roman Moiseev is a lecturer in pharmaceutics at the University of Reading. Jon Pugh is senior lecturer for applied drama: education, well-being, community at the University of Wales Trinity Saint David. Geyen Sasha Surendran is lecturer in anatomy and physiology at British International University. All are part of the Active Learning Network.
If you would like advice and insight from academics and university staff delivered direct to your inbox each week, sign up for the Campus newsletter.
