South African College of Applied Psychology

In my view, the success of the first Global Dialogue on AI Governance would lie in its ability to move beyond high-level discussions and produce outcomes that are grounded in real-world practice, particularly within education systems. From my experience working in higher education and curriculum development, one key outcome would be the development of guiding principles that are not only human-centred in theory, but practically applicable in teaching and learning environments. AI is already shaping how knowledge is created, assessed, and experienced, and governance frameworks need to reflect these shifts in meaningful ways. A second important outcome would be the genuine inclusion of Global South perspectives. Working within a South African context, I see firsthand how factors such as access, infrastructure, and institutional capacity influence how AI is adopted. If these realities are not embedded into governance conversations, there is a risk of deepening existing inequalities. Additionally, I believe success would involve recognising the psychosocial dimensions of AI. As someone whose research explores the human experience of AI in education, I see the need for governance approaches that consider how AI impacts cognition, identity, agency, and the student learning experience. Finally, the Dialogue should establish ongoing, collaborative structures that connect policy with practice. This would ensure that governance remains responsive, inclusive, and informed by those working directly with AI in everyday contexts. Ultimately, success would mean creating a foundation for AI governance that is not only globally relevant, but contextually aware, interdisciplinary, and grounded in lived experience.

• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Protection and promotion of human rights
• Transparency, accountability, and human oversight

My selected priorities reflect my work at the intersection of artificial intelligence, higher education, and the psychosocial dimensions of teaching and learning. AI capacity-building is a key priority, as I am directly involved in curriculum development and the integration of AI into higher education systems. There is an urgent need to equip both educators and students with the knowledge and competencies required to engage with AI critically and effectively, particularly within resource-constrained contexts. The social, economic, ethical, and cultural implications of AI are central to my research, which explores the psychosocial impact of AI in education. I am particularly interested in how AI influences cognition, agency, identity, and the overall student learning experience, and believe these human-centred considerations must inform governance approaches. The protection and promotion of human rights is closely linked to my focus on equity and inclusion within Global South education systems. Issues such as access, infrastructure, and digital inequality significantly shape how AI is adopted and experienced, and must be addressed to avoid reinforcing existing disparities. Finally, transparency, accountability, and human oversight are critical in ensuring responsible AI use in educational contexts. As AI becomes embedded in assessment, content generation, and decision-making processes, clear frameworks are needed to maintain academic integrity, ethical standards, and trust. Together, these priorities reflect the need for AI governance that is practical, inclusive, and grounded in lived educational realities.

One key cross-cutting issue that is not fully captured in the listed themes is the evolving relationship between humans and AI as collaborative partners, particularly in knowledge creation and learning. In higher education, AI is no longer simply a tool, but is increasingly shaping how students think, write, and engage with knowledge. This raises important questions about authorship, agency, and the development of critical thinking skills, which extend beyond existing ethical or technical considerations. A second emerging issue is the need for context-sensitive implementation frameworks. While many governance discussions focus on universal principles, there is often insufficient attention given to how these principles translate into practice across diverse institutional and socio-economic contexts. In Global South settings, factors such as infrastructure limitations, varying levels of digital literacy, and institutional readiness significantly influence AI adoption. Governance approaches that do not account for these differences risk being ineffective or exclusionary. Additionally, there is a growing need to address the impact of AI on assessment and academic integrity. Traditional models of evaluation are being challenged, and there is a lack of clear, globally informed guidance on how to redesign assessment in ways that remain valid, fair, and aligned with learning outcomes in an AI-integrated environment. Finally, the pace of AI development highlights the need for adaptive governance mechanisms. Static regulatory approaches may quickly become outdated, making it essential to develop flexible, iterative frameworks that can respond to rapid technological and societal change. These issues underscore the importance of governance approaches that are not only technically robust, but also pedagogically informed, contextually grounded, and responsive to evolving human-AI dynamics.

Within the South African higher education sector, governance gaps in AI are creating both significant challenges and emerging opportunities. One of the most pressing challenges relates to uneven AI capacity and infrastructure. While some institutions are beginning to integrate AI into teaching, learning, and assessment, others face constraints such as limited digital access, resource shortages, and varying levels of staff preparedness. This creates disparities in how students experience and benefit from AI and may reinforce existing inequalities across the sector. A second challenge is the lack of clear, contextually relevant governance frameworks. Many institutions are navigating AI integration without comprehensive policies on ethical use, academic integrity, data privacy, and accountability. This has led to uncertainty among educators and students, particularly regarding appropriate use of AI in assessments and knowledge production. At the same time, these gaps present important opportunities. AI has the potential to enhance curriculum innovation, personalise learning, and expand access to educational resources, especially in constrained systems. In my work, I have seen growing interest in embedding AI into curriculum design in ways that support critical thinking and student engagement. There is also an opportunity to develop governance approaches grounded in Global South realities. By addressing local challenges such as access, language diversity, and institutional capacity, South Africa can contribute context-sensitive models for responsible AI integration. Ultimately, while governance gaps create uncertainty and risk, they also open space for innovation and more inclusive AI practices in higher education.

The AI Dialogue can play a critical role in advancing international cooperation by serving as a bridge between diverse national, regional, and sectoral perspectives on AI governance. In a rapidly evolving landscape, no single country or institution can effectively govern AI in isolation, making coordinated and inclusive dialogue essential. One key role of the Dialogue is to facilitate the development of shared understandings and common principles that are flexible enough to be adapted across different contexts. This is particularly important for ensuring that governance approaches are not dominated by a narrow set of technological or geopolitical interests, but instead reflect a broader range of lived realities, including those of the Global South. The Dialogue can also support capacity-building by enabling knowledge exchange between countries and sectors. For example, higher education institutions navigating AI integration can benefit from shared practices, policy frameworks, and lessons learned across regions. This type of collaboration can help reduce disparities in AI readiness and support more equitable participation in the global AI ecosystem. In addition, the Dialogue can promote interdisciplinary engagement by bringing together stakeholders from education, social sciences, policy, and technology. This is essential for developing governance approaches that account for the human, social, and ethical dimensions of AI, rather than focusing solely on technical considerations.

The AI Dialogue should build on existing global and regional initiatives that are already shaping AI governance, while strengthening coordination between them. Key examples include UNESCO's Recommendation on the Ethics of Artificial Intelligence, which provides a comprehensive human rights-based framework, and the OECD AI Principles, which have informed many national strategies. In the African context, initiatives such as the African Union's Continental AI Strategy and emerging national AI policies offer important region-specific perspectives that should be meaningfully integrated. In the higher education sector, there are also growing institutional and network-based efforts focused on AI integration, academic integrity, and digital transformation. These include university-led frameworks, professional bodies, and collaborative research initiatives that are actively responding to the practical challenges of AI adoption in teaching and learning. The added value of the AI Dialogue lies in its ability to connect these often fragmented efforts into a more coherent and inclusive global ecosystem. It can serve as a platform that brings together policy, practice, and research, ensuring that governance approaches are informed by real-world implementation across sectors. Importantly, the Dialogue can amplify underrepresented voices, particularly from the Global South, and ensure that existing frameworks are not simply adopted, but critically adapted to different socio-economic and institutional contexts. It can also facilitate knowledge exchange by creating structured opportunities for stakeholders to share lessons, challenges, and innovations. The AI Dialogue can provide continuity by linking existing initiatives into an ongoing, collaborative process rather than isolated efforts. In doing so, it can help move global AI governance from parallel developments towards more aligned, contextually relevant, and actionable approaches.

Different stakeholders can contribute to the AI Dialogue by bringing complementary forms of expertise, experience, and accountability. Governments can provide regulatory direction and align national strategies with global principles. Academia can contribute research, critical perspectives, and evidence-based insights, particularly on the social and educational implications of AI. Industry can offer technical knowledge and practical implementation experience, while civil society can ensure that human rights, equity, and public interest considerations remain central. Importantly, practitioners such as educators should be included to reflect how AI is experienced in everyday contexts. To support meaningful participation, the format of the AI Dialogue should be multi-layered and inclusive. This could include a combination of high-level plenary sessions, thematic working groups, and open consultation processes. Thematic groups should be interdisciplinary and focused on specific areas such as education, ethics, or capacity-building, allowing for more detailed and practice-oriented discussions. The Dialogue should also incorporate regional consultations to ensure that local contexts and priorities are meaningfully represented. This is particularly important for including perspectives from the Global South, where challenges and opportunities may differ significantly. In terms of structure, the Dialogue should not be a one-time event but an ongoing process with clear mechanisms for continuity. This could include regular reporting, shared knowledge platforms, and iterative feedback loops that allow stakeholders to engage over time.

Global discussions on AI governance continue to underrepresent voices from the Global South, particularly those working within resource-constrained systems such as public higher education. While AI development is often driven by well-resourced contexts, its impacts are global, and governance frameworks do not always reflect the realities of regions facing challenges related to infrastructure, access, and digital inequality. In addition, there is limited representation from practitioners who engage with AI in everyday contexts. Educators, for example, are actively navigating the integration of AI into teaching, learning, and assessment, yet their insights are often absent from high-level policy discussions. Similarly, students, whose learning experiences are directly shaped by AI, are rarely included in governance conversations. Interdisciplinary perspectives are also underrepresented. Much of the discourse is dominated by technical and policy expertise, with less input from fields such as psychology, education, and the social sciences. This limits understanding of the broader human and societal implications of AI, including its impact on cognition, identity, and behaviour. To address these gaps, more inclusive and accessible participation mechanisms are needed. This could include targeted regional consultations, partnerships with educational institutions, and structured opportunities for practitioner input. Virtual engagement platforms can also help reduce barriers to participation. In addition, deliberate efforts should be made to include diverse disciplinary perspectives through interdisciplinary working groups. Providing support for participation, such as funding, capacity-building, or institutional partnerships, would further enable meaningful inclusion.

To foster meaningful and dynamic engagement, the AI Dialogue should move beyond traditional panel-based formats and adopt more interactive, practice-oriented approaches. One effective format would be scenario-based workshops, where participants engage with real-world case studies, such as the use of AI in education, healthcare, or public services. This would allow stakeholders to collaboratively explore governance challenges and develop context-sensitive solutions, rather than remaining at the level of abstract discussion. Another valuable approach would be multi-stakeholder roundtables that intentionally bring together diverse participants, including policymakers, educators, technologists, and civil society actors. Structuring these sessions around specific problem statements can encourage focused, outcome-driven dialogue and ensure that different perspectives are meaningfully integrated. Digital and hybrid engagement platforms are also essential. Interactive online spaces, such as moderated forums or collaborative workspaces, can enable ongoing participation beyond formal sessions, particularly for stakeholders who are unable to attend in person. This supports more inclusive and continuous engagement.

Several existing policies and practices offer valuable approaches to effective AI governance. At a global level, UNESCO's Recommendation on the Ethics of Artificial Intelligence provides a comprehensive human rights-based framework that emphasises inclusivity, accountability, and ethical design. Similarly, the OECD AI Principles have guided national strategies by promoting transparency, robustness, and responsible innovation. In practice, higher education institutions are beginning to develop context-specific frameworks for AI use in teaching and learning. These include institutional guidelines on academic integrity, ethical AI use in assessment, and the redesign of curricula to incorporate AI literacy. Such approaches demonstrate how governance can be embedded within everyday educational practice rather than treated as a purely regulatory function. Emerging practices in assessment redesign are also particularly relevant. Some institutions are shifting towards authentic, process-based assessments that emphasise critical thinking, reflection, and application, reducing over-reliance on traditional outputs that can be easily generated by AI. This represents a practical response to the challenges AI poses to academic integrity. Digital platforms that support transparency and accountability are another important development. For example, tools that provide AI disclosure mechanisms or track AI-assisted contributions can help maintain trust while acknowledging the evolving role of AI in knowledge production. Regional efforts such as the African Union's Continental AI Strategy highlight the importance of context-sensitive governance that addresses issues of access, capacity-building, and equitable participation.