New research findings on use of conversational agents in higher education

A new publication from the Commonwealth of Learning (COL), authored by Professor Sara de Freitas and Professor Denise Whitelock at The Open University's (OU) Institute of Educational Technology, offers timely guidance on the responsible use of conversational agents in tertiary education. Conversational Agents in Education: Context and Case Studies of Practice in Tertiary Education traces the evolution of these AI enabled systems, examines their pedagogic and operational roles, and presents real world case studies from UK institutions.

Reflecting on the publication, Professor Denise Whitelock said: “What I value most about this collection is how honestly it reflects the day to day realities of teaching with AI. These practitioners are tackling real challenges with creativity and compassion, and their insights have inspired my own thinking. I hope this publication encourages others to explore AI in ways that strengthen, rather than replace, the human relationships at the heart of education.”

Conversational agents: long-standing tools with renewed relevance

Although generative AI has rapidly become part of everyday life in higher education — from students using it to draft assignments to staff using it to design learning materials — the roots of conversational agents stretch back more than 50 years. Early expert systems and hypermedia prototypes laid the groundwork for today’s text, voice and avatar based assistants that learn from data and human interaction.

When used responsibly, conversational agents can reduce administrative burdens, personalise support, improve feedback, and help students access timely assistance. Yet these benefits must be balanced against risks such as inaccuracy, bias and over reliance on automated systems. The report emphasises the need for human oversight, staff training and regular evaluation to ensure ethical and informed use.

Evidence-based guidance for a rapidly evolving landscape

Drawing on decades of Artificial Intelligence in Education (AIED) research, Professors de Freitas and Whitelock analyse how conversational agents have developed and how they are currently being deployed across UK higher education. The publication includes case studies ranging from simple text based assistants to more advanced systems using animated avatars, illustrating how institutions are using these tools to enhance learning guidance, boost student engagement and streamline information retrieval.

Importantly, the report demonstrates that conversational agents can be implemented effectively even in resource constrained environments. This aligns with COL’s vision for Frugal AI — open, affordable and lightweight systems that can run locally, adapt to diverse contexts and support digital sovereignty. Unlike large proprietary models requiring extensive cloud infrastructure, conversational agents can be built using modular or open source frameworks and deployed in low connectivity settings.

Supporting ethical, equitable and human centred AI adoption

The publication also reflects the principles of the Gaborone Statement, which calls for equitable and human centred approaches to AI. By showing how conversational agents can augment rather than replace educators, the report highlights their potential to provide personalised support without costly infrastructure, while remaining transparent, adaptable and locally governed.

Far from being futuristic technologies, conversational agents are already well established tools that can help institutions across the Commonwealth strengthen teaching, learning and student support. This evidence based resource aims to support educators, leaders and policymakers as they navigate responsible and inclusive AI adoption.

Professor Denise Whitelock is a former Chair of COL.