by Akweley Ohui Otoo and Kate O’Riordan

As part of a reciprocal mentoring project with colleagues at the University of Ghana we have a reflection on cross cultural experiences of AI in education.

Introduction

Generative AI has a longer history, but 2022 marked the popularisation of generative AI, when user interfaces were made widely available. Take up of generative AI has exponentially increased since then, the implications and impacts of which are still emerging. However, some things are clear. Like the web moment of the 1990s, the web 2.0 social media moment of the early 2000s, and the pandemic driven adoption of technology platforms across new areas, this is an exponential expansion that will continue to have major shaping effects, even when there is some fall off and retraction. 

In the UK, public discussion about generative AI moved very quickly into education, particularly higher education. Media coverage of major news outlets in the UK in 2023 show a very high incidence of references to university students. Higher Education is not always mainstream news in the UK, but generative AI discussions in the news brought discussions of HE into mainstream coverage. These usually focused on assessment, and the debate was characterised by concerns about students cheating and assessment design and, less frequently, about AI marking. In Ghana, public discussion was triggered even earlier, particularly when Google AI opened in Accra in 2018. AI has been central to government and business discourse about innovation and economic growth, and generative AI has been central to debates about education in the same period. 

In this context HE institutions, academic bodies, technology providers and advocates, in both Ghana and the UK, and internationally, have moved to look at approaches, principles and scenarios, to shape academic practice in a post generative AI context. They have also looked at new user interfaces, tools and technologies, exploring innovation and use through both boosterish and promotional discourses, and highly critical and resistant ones. This utopian/dystopian dynamic characterises discussions, imaginaries and implementation of new technologies and has been explored in the social studies of science literature. This is usually tempered through the language of challenges and opportunities. 

In this broader context, we compare the conversations, policies and technologies in two different university contexts: Ghana and Sussex. The University of Ghana and the University of Sussex have a long history of connection, and an important strategic partnership. This has historically been focused on research collaborations in the sciences, education and international development. Through 2023-2025 this also included a project through CEGENSA (led by Professor Deborah Atobrah) at the University of Ghana, and the gender equality workstream in the EDI unit at the University of Sussex (led by Professor Sarah Guthrie). This project established a reciprocal cross-cultural mentoring scheme across the two universities, which is the context in which these reflections about AI have been developed. 

More broadly Sussex and Ghana have had strong education relationships nationally, through students and alumni, and regional connections including the Fiankoma project in the 1990s (Pryor, 2008). This was another cross-cultural project which aimed to link the community of Fiankoma in Ghana with people and educational institutions in Sussex (in the UK) through digital technologies. Educators and students in both settings produced accounts of their lives using digital media that were turned into a web site for cultural exchange and development education. 

The current Ghana-Sussex cross cultural, reciprocal mentoring project paired colleagues across both institutions, from different disciplines, and across different roles. When we first met, for the authors of this paper, we were coming from different disciplinary backgrounds. Otoo is a social psychologist in the Department of Distance Education and O’Riordan is a digital media scholar, from media and communications, and currently based in the Vice Chancellor’s Office. Our respective institutions were in very different places in relation to debates about generative AI, and at the same time there were very strong common themes. We outline and reflect on these differences and similarities here. 

Post pandemic

The COVID-19 context informs what has happened with generative AI and this played out differently in relation to the two institutions in relation to educational delivery. At the University of Ghana, there was already a very strong online teaching offer (ODL) for the discipline of education. This meant that staff, and students were already used to using a learning management system (LMS) called Sakai as the educational context, and had been using this for over a decade. ODL at the University of Ghana is largely intended for Ghanian students, and aligned with teaching methods in the on campus offer in many ways. The Sakai support unit was based in their department. However, for the on campus offer in the rest of the university there was no use of a LMS, and there was very little experience of this. In the context of the pandemic and the shift to remote education for everyone, the rest of the university started using the same platform that had already been in use, but previously only for one area. As a consequence of this shift, Sakai became the university platform and the support unit is now based in the main campus Balm Library. 

At the University of Sussex, conversely, a LMS called Canvas was already used throughout the university to support the on campus offer. There was also a pre-pandemic online distance learning (ODL) offer as part of the overall educational offer, which also used Canvas. ODL at Sussex is very distinct from the in-person campus offer, and is delivered to students in other parts of the world and is largely nonsynchronous. However, the on campus offer, which was the main educational offer, was already mediated through Canvas as standard, and LNS use had been in play for decades at this point. In the pandemic the same platform that everyone was already using, was used more intensely, and in a remote mode for all teaching. This was often a synchronous offer that replaced timetabled teaching with online sessions. Although there was some unevenness of experience and expertise, and additional platforms were also brought to bear, this story was more about intensification of the use of the LMS rather than a wholesale change of practice. 

Therefore, there was a contrast between the Sussex and Ghana experiences in the pandemic technology adoption for education. At the University of Ghana, a smaller group of staff and students with specific expertise saw wholesale adoption of the previously ODL-only platform by the rest of the staff. At the University of Sussex, there was wholesale intensification of an existing platform and set of technologies that had already been in widespread use across the offer. 

Building on the contrast between the wholesale adoption of the Sakai platform by the University of Ghana and the intensification of existing platforms at the University of Sussex, there was also an added implication for the colleagues at the University of Ghana, with existing expertise in use of the Sakai platform.

The distinction is that at the University of Ghana, the rapid, wholesale adoption of the platform placed an immense and sudden workload on the smaller group of staff and ITprofessionals with the existing expertise. They instantly transitioned from being expert users and platform managers to becoming the institution’s primary, often overburdened, trainers,first-line support staff, and technical consultants for the entire faculty and student body. Their expertise became a critical, non-negotiable bottleneck to the institution’s operational continuity.

Conversely, at the University of Sussex, where technologies were already in widespread use, the intensification meant (at least in theory) the existing experts could focus more on scaling, optimizing, andproviding advanced pedagogical support, rather than on fundamental adoption and crisis-level initial training. However, in practice, the pandemic revealed varying levels of expertise with the existing system, despite its ubiquity, and there was also an uneven impact, with some patterns of over burdening and bottlenecks of expertise.

In 2024 the University of Ghana celebrated the 10th Anniversary of the Sakai Learning Management System (LMS) with a two-day blended conference. Professor Yaw Oheneba-Sakyi pioneered the introduction of Sakai at the university. This Learning Management System proved to be a lifeline for the entire university community—including teaching staff and students—during the COVID-19 era, leading to its wholesale adoption by everyone.

One common theme ran through the submissions and speeches of presenters at the conference which was that the University should embrace and integrate new technologies in its teaching and learning. Prof. Ohene Sayki detailed the system’s success in advancing the University’s academic mission, specifically citing the establishment of the MA in Distance Education and E-Learning, its successful integration into PhD teaching, and the global leadership roles attained by its alumni.  He outlined a strategic path forward focused on expanding mobile access, the responsible adoption of artificial intelligence and deepening in collaboration.

At the University of Sussex, Canvas was introduced as a VLE in 2018, replacing a previous system (StudyDirect). It was introduced and supported by the Technology Enhanced Learning group. Support for Canvas is delivered by the Educational Enhancement team and is reviewed in an ongoing way, with rolling workshops and resources. There are templates for use, and resources for good practice. When it is used for nonsynchronous ODL, there is a different approach to the creation of interactive learning materials, supported by a partner specialist (currently Boundless). Good practice in the use of Canvas is celebrated in teaching and learning workshops and conferences at Sussex, but the platform itself hasn’t been the focus of a university conference. 

Post AI (student and staff discussions and practices after AI)

At the University of Ghana, in 2024 the university updated its approach to plagiarism by including the use of AI in its approach to misconduct. However, there has also been a paradigm shift in how AI is viewed in recent times. It has shifted from the perspective of AI as a tool for cheating to that of a tool used in the efficiency of knowledge production and an assistance to both staff and students in teaching and learning pedagogy. A recent study with 17 PhD students in the College of Education of the University of Ghana was conducted to explore the use of generative AI tools in their work. The students explored several Generative AI tools in their various capacities.

Findings from this demonstrate that students used a variety of GenAI tools across their academic tasks:

• For brainstorming and generating initial written content. ChatGPT and Copilot were adopted.  
• For language improvement and editing academic writing, Grammarly and QuillBot come in handy and for clearer visual aids and presentations, the DALL.E was used.

The project also explored the pros and cons of using these Gen. AI tools in academic writing.  There were positive impacts that translated into increased academic confidence, improved time efficiency, fostered self-directed learning and improved digital literacy.  On the other hand there are emerging concerns about metacognitive laziness, diminished qualitative interpretation and risks to intellectual agency and overreliance and “Al-holic” phenomenon. The study concluded that GenAI can empower experiential learning only when integrated with pedagogical intent and ethical awareness.

At the University of Ghana, AI use has become normalised and is used in everyday operations. For example, the use of AI notetaker in remote meetings, and the use of AI add-ons in the LMS and other IT systems. However, the University has yet to adopt formal principles or policy beyond the updated plagiarism policy. At the same time, the University is a centre for research into AI in schools, and education more broadly. For example, the College of Education held a virtual panel discussion centred around the theme, “Generative AI: African Perspectives on its Challenges and Prospects.” at the 2024 Day of Scientific Renaissance of Africa (DSRA). 

At the University of Sussex students and staff engaged with generative AI, but there was (and continues to be) a lot of uncertainty about the legitimacy of using AI. In the UK HE sector, a number of AI surveys, reports and policy notes, have been (and continue to be) produced, through actors such as the Higher Education Policy Institute (HEPI) and the Joint Information Systems Committee (JISC). A group of universities in the UK developed the Russell Group Principles and many institutions have signed up to these.

At the University of Sussex, a Community of Practice (CoP AI) was initiated by the Educational Enhancement team (EE). The first change to policy and guidance at Sussex was the amendment of academic misconduct to include generative AI in concepts of plagiarism, personation and fabrication (existing integrity concepts). Alongside this, guidance for staff and students, and a number of case studies were shared. In 2024 there was a university-wide engagement and an AI summit that culminated in the development of a set of institutional principles.

At Sussex, during the university-wide engagement, concerns and ideas were raised across integrity, the meaning of knowledge, automation, legitimacy and environmental impact. The principles themselves focused on seven key areas to:

• build on Sussex’s world leading research in AI by investing in ongoing, interdisciplinary research on AI in education
• develop strong digital capability and critical AI literacies for our students and staff
• deepen ethical standards
• protect our academic integrity and student experience
• foreground accessibility and inclusion in our approach to the use of AI in education
• safeguard our community against malicious or illegal use of AI
• commit to clearly communicated and transparent governance

The last point, about transparency, has already become very difficult to deliver on, because the underpinning systems across the university now, increasingly, have AI functionality built in. This is not always visible, and software updates are included as an automatic default. For example, the lecture capture platform had an automated generative AI captioning function built into the latest upgrade. Our capacity as an institution to clearly communicate how and where AI is already in our systems, and to be transparent about the governance of this, is already significantly limited by the technical design.

Both universities are continuing to build up resources, guidance and training, both in-house and with external partners.

Conclusion

At the time of writing, the University of Sussex continues to build on the AI principles. This includes staff training, and resources, and continuing to identify opportunities to support staff and students to develop capability and critical literacies. There has been a return to in-person, invigilated exams, and many subject areas are rethinking their approach to assessment post-AI. This includes pilots for proctoring software and lockdown browsers. Responding to the challenges and opportunities through these technological disruptions is work that is ongoing and has to be made and remade. The AI landscape continues to change rapidly, and understanding, and access is uneven across the University. A broader set of principles that apply to the operational and research dimensions of the University is also necessary and is being led through the Digital and Data Task Force as part of the strategy, Sussex 2035. 

The University of Ghana is developing an AI policy, and is actively researching the impact of AI in education. For example, Dr Freda Osei Sefa from the College of Education is leading research on the use of AI in basic (primary and lower secondary) schools. Recently, the Kwame Nkrumah University of Science and Technology (KNUST) has introduced a compulsory module for all students, and this is in the pipeline for the University of Ghana. This is largely focused on careers and future job prospects. A concern shared at the University of Sussex. 

It is clear that generative AI particularly, and AI more generally, is an important feature globally and this plays out in both universities in different ways but with very strong common themes.     

References

Acquah, R (2024) ‘University of Ghana revises plagiarism policy to include AI’

https://www.myjoyonline.com/university-of-ghana-revises-plagiarism-policy-to-include-ai/ Source: Raymond Acquah, 26 February 2024 10:43am

Pryor, J (2008) Analysing a Rural Community’s Reception of ICT in Ghana, in Van Slyke, C (ed) Information Communication Technologies: Concepts, Methodologies, Tools, and Applications.