How AI is designed and integrated into education can have different consequences for different communities. Whose needs, experiences and perspectives shape its use, who benefits and whose interests risk being overlooked? In this article, Prof. John Traxler reflects on the latest AI and Education hackathon, organised by the University of Manchester on behalf of UNESCO, which brought together educators, managers and developers from across South Asia to explore AI, design justice and its implications for education.
AI and Education: Insights from the UNESCO South Asia Hackathon
Author: Prof. John Traxler, UNESCO Chair, Commonwealth of Learning Chair and Academic Director of the Avallain Lab
St. Gallen, October 29, 2026 – For the past four years, the Digital Technology, Communication and Education (DTCE) group at the University of Manchester has organised a ‘hackathon’. On behalf of UNESCO, this series of webinars, discussions, activities and group projects explored the theme of ‘AI and Education’.
In 2026, the hackathon brought together over 100 participants from the public and private sectors, including educators, managers and developers, from across South Asia.
One of the hackathon’s underlying aims was to promote ‘design justice’ in AI and education by enabling participants to examine, challenge and confront how AI in education reinforces power structures, systemic inequality and the marginalisation of communities, cultures and countries, including everyday practices within schools, colleges and universities.
Understanding Design Justice
Design justice is an approach that examines how technologies and systems can reinforce or challenge existing inequalities. It recognises that decisions about how technologies are designed, developed and implemented are not neutral, and that these decisions can have different consequences for different communities.
Design justice is also a response to the perceived economic attractions of scale delivered by global digital technologies, to the distractions of ‘localisation’ that seem like a technical fix for ‘one-size-fits-all’ technologies and to the ways these threaten less powerful or technologically savvy communities and cultures outside the global mainstream. Design justice, as the name suggests, is an ethical response to an ethical threat.
Central to this approach is involving those most affected by a technology in shaping its development. In education, this means considering whose needs, experiences and perspectives inform the use of AI, who benefits from it and whose interests might otherwise be overlooked.
Putting Design Justice into Practice
As part of the hackathon, participants worked in groups. They were assigned projects to produce and present end-of-course artefacts, for example, posters, reports, presentations or videos. Each group was supported by a university mentor and tasked with analysing the needs of a specific group or community, for example, teachers or parents in their own country or culture, and producing something useful for that group or community.
This demonstrated how design justice could be put into practice, not as an alternative universal prescription but as methods and tools that could be adopted and adapted. In that respect, the hackathon demonstrated the processes, not the products. Putting design justice into practice around educational AI is not just timely but uniquely challenging. Not only is most of the data AI trains on skewed and biased, away from the values, languages, ideas, knowledge and experiences of most of the world’s communities and cultures, but so is the way AI processes and ‘reasons’ this data.
For example, one group presentation followed the process from engaging with their target audience, university students, to designing a series of posters that met the needs they uncovered. Another approach was to build a teaching resource or lesson plan based on the reported problems with AI use. Both show local resources growing out of local needs, in spite of the easy access to distractions of web-enabled, mobile-friendly but biased chatbots such as Claude and Gemini.
These artefacts were often very accessible, colourful and visual. Many also reflected a cautious or critical approach, rather than advocating for an uncritical adoption or exploitation of AI in education. There was recurring emphasis on and determination to preserve human agency and engagement wherever AI was used.
The artefacts took different approaches. Some were practical, aimed at students, teachers and parents in their day-to-day lives; others were more critical, drawing attention to the harm and bias associated with overly naive or optimistic AI adoption.
Participants worked remotely from their respective countries, collaborating across multiple time zones while navigating poor connectivity and the demands of their jobs and studies. Mixing the group composition was a tactic to foster cross-cultural collaboration and understanding, but perhaps a more purist approach would be to have homogeneous groups, each drawn from one specific culture or community.
Exploring Key Hackathon Artefacts
Some of the most visual artefacts included this enchanting sequence of cartoons helping Indian parents, at their local chai stall, learn about the AI their children might use: firstly, the parents gather for chai; then, as a consequence, provide help with homework, warn about sharing personal information and get encouraged to stay involved; and then finally, gather back at the stall to recap. This was based on findings within the community.
The cartoons illustrate how familiar cultural settings and everyday experiences can make discussions about AI more relevant and accessible. Rather than approaching AI literacy as a purely technical subject, the resource connects it to parents’ existing roles in supporting their children’s education.
Notions of literacy, recently digital literacy and now increasingly AI literacy, have a substantial and somewhat controversial history, dating back at least into the UK literacy campaigns of the 1970s. The issues and tensions, however, remain the same, as much in India now as in the UK then. Is literacy about skills, practices, and information, about know-how, necessary to use the medium competently, to engage with reading the written page, the computer screen or the AI chatbot, or is it about criticality, about questioning the page, the screen and the chatbot, about know-why? These tensions have always had a political dimension, recycled through the successive technologies, because they potentially challenge what is being taught, what is being consumed and what is being accepted, same then, same now.
The current cartoons not only make these points in terms of familiar cultural settings and everyday experiences but should remind us that those settings and experiences are not universally comparable or equivalent.
The need for and nature of AI literacy, especially the kind of critical AI literacy that encourages people to question and challenge answers, varies across cultures and contexts. It is also worth considering how the values and ideas of global AI corporations may influence approaches to AI literacy, and whether these always reflect the needs and priorities of local communities.
Meanwhile, in another video, a student under pressure to submit an assignment blindly trusts AI and needs guidance to use it more responsibly and critically.
Other artefacts used vox-pop sequences with a range of people in their country, Bangladesh, for example. They discussed their experiences and attitudes to AI in their lives and their work, illustrating how pervasive and accessible AI has become in parts of South Asia.
This group of posters come with a toolkit for teachers; the emphasis here is on teachers and learners recognising that AI must support human agency, not replace it. Another set of posters, entitled ‘AI for teachers’, supports AI choice, discusses AI in traditional teaching, and covers creating lesson plans and tests while staying aware of security and ethics.
This presentation takes a more critical stance on cultural bias in AI, especially for teachers and learners in Asia. The team raised important questions about whose languages, cultures and perspectives are represented in AI systems and how potential biases might influence the educational content and information these technologies produce.
Reflection and Outcomes
Looking back on the hackathon and reflecting on participant feedback suggests it was a hugely successful attempt to help a very disparate group work together, think through the deeper implications of AI for education in their communities, and show that this can lead to practical, tangible outcomes.
Beyond the artefacts themselves, the hackathon raises wider questions about whose experiences and perspectives shape AI in education, how communities can participate meaningfully in its development and how educators can preserve human agency as these technologies become more widespread.
Institutional Context and Bias
We must also acknowledge that the hackathon was organised within an institutional context that brings its own perspectives and bias to discussions about AI and education. As a university based in the Global North, the University of Manchester operates within particular academic and institutional traditions. Even if a university in the Global South hosted the hackathon, it could still be argued that it buys into a specific globalised version of higher education that might not align with the values and ideals of different local cultures, communities, castes and classes.
UNESCO, too, operates within an institutional framework that serves the interests of its Member States, whose governments play a central role in determining its policies and priorities. So who is truly being represented, and do the discussions about AI in education give sufficient attention to learning beyond the formal national education systems? UNESCO publications often present learning as something that happens only within these systems. This means groups of people, often women, girls and minority groups, are ignored, overlooked, disadvantaged or failed by education systems or that people whose learning takes place informally beyond the reach of teachers are also overlooked.
In terms of digital learning, much learning through personal technologies and social media may take place without sufficient guidance or critical engagement. Similar concerns now arise with AI, as its widespread use raises questions about how to support individuals to engage with these technologies thoughtfully and responsibly.
The hackathon has made a contribution to these discussions, and its message can continue to inform wider thinking about AI and education.
Design Justice and the EdTech Industry
It might also be the case that design justice is treated as a fringe luxury, rather like organic vegetables, workers’ cooperatives and open-source systems, in the face of established global systems, methods and practices, and in the case of design justice in educational AI, we must surely ask about the relationships between design justice and the edtech industry. Might there be synergy and overlap; can either side learn from the other? Could the edtech industry acknowledge and incorporate local perspectives without merely appropriating and exploiting them? Even with an authentic willingness, this will be a challenge, given that the edtech industry is downstream of massive, unresponsive AI corporations over which it has little influence.
While design justice is possible and while this hackathon focused on communities in South Asia, why not also consider how these principles can be applied to other communities whose perspectives may be overlooked, including the neurodiverse, the indigenous, the homeless and the Traveller communities?
Supporting the hackathon was part of Avallain Lab’s ongoing work exploring the ethical, cultural and pedagogical implications of AI across diverse educational contexts, in collaboration with leading academics and universities. These insights also contribute to the broader research and ethical foundations of Avallain Intelligence, Avallain’s framework for the responsible use of AI in education, which prioritises human agency, fairness and the safe, inclusive use of educational technology.
The Background to the Hackathon
There were 120 participants from 11 institutions: University of Manchester; University of Liberal Arts Bangladesh; University of Dhaka, Bangladesh; University of Kathmandu, Nepal; Key Academy, Nigeria; Mahidol University, Thailand; Universitas Kristen Satya Wacana, Indonesia; Maldives National University, Maldives; Royal University of Bhutan, Bhutan; University of Colombo, Sri Lanka; Symbiosis Skills and Professional University, India.
In this year’s iteration, the organisers intentionally encouraged interdisciplinary cohorts where possible, drawing from education, social sciences, humanities, policy or technology-related programmes.
UNESCO South Asia supported collaboration between participating universities, and each institution identified focal-point colleagues who supported recruitment, onboarding, attendance and communication with participants locally. Dr Taslima Ivy, Mehdi Hassan and Martyn Edwards organised successive hackathons, assisted by Jian Xi Teng: Programme Specialist (Education) for UNESCO South Asia, Norihide Furukawa: Head of Education, UNESCO Dhaka, and Gazi Fabiha Noshin Meem, Education Consultant at UNESCO Dhaka, many of whom were involved in a parallel seminar.
The expert and scholarly contributions were as follows:
‘Frameworks for AI Literacies: A Comparative Analysis of UNESCO’ and ‘The OECD Literacies: A Comparative Analysis of UNESCO and the OECD’ from Susan Robertson, Professor of Sociology of Education at the University of Manchester, and Dr Michele Martini, Postdoctoral Fellow at the Faculty of Communication, Culture and Society at the Università della Svizzera italiana (USI)
‘Against the Grain of Corporate Hype: Strategies for Critical and Grounded Teaching about the Impact of Generative AI on Learning, the Environment and Society’ from Eleanor Dare of Cambridge Digital Humanities, University of Cambridge
‘Transformative learning from a sustainability perspective: An essential fundament to AI use and competencies?’ from Amanda Banks Gatenby and Susan Brown, University of Manchester, Alex Dunnedin from the Ragged University
‘Assessment: what is it good for?’ followed by ‘Putting Enjoyment Into Assessment: Or How I Learned To Stop Worrying And Live With AI’ from Drew Whitworth and Martyn Edwards, University of Manchester
‘Dialogic Educational Futures: Foundations for Critical Design in an Uncertain AI Landscape’ from Louis Major, University of Manchester
‘AI literacies and Global Justice’ from Helen Beetham with Duncan Ivison, Vice Chancellor, University of Manchester
‘Ensuring Inclusivity and Equality in AI Competency Building: Insights from Global Higher Education Contexts’ from Skye Zhao, University of Manchester, Fengchun Miao, Chief, Unit for Technology and AI in Education, UNESCO and Melti Oktavianda, Manchester Institute of Education, University of Manchester
‘Decolonising Educational AI’ from me, Prof. John Traxler, plus the chance to serve as discussant for most sessions.
About Avallain
For more than two decades, Avallain has enabled publishers, institutions and educators to create and deliver world-class digital education products and programmes. Our award-winning solutions include Avallain Author, an AI-powered authoring tool, Avallain Magnet, a peerless LMS with integrated AI, and TeacherMatic, a ready-to-use AI toolkit created for and refined by educators.
Our technology meets the highest standards with accessibility and human-centred design at its core. Through Avallain Intelligence, our framework for the responsible use of AI in education, we empower our clients to unlock AI’s full potential, applied ethically and safely. Avallain is ISO/IEC 27001:2022 and SOC 2 Type 2 certified and a participant in the United Nations Global Compact.
Find out more at avallain.com
_
Contact:
Daniel Seuling
VP Client Relations & Marketing
