Nigeria Innovation Summit

For Africa, the success of the first Global Dialogue on AI Governance will not be measured by the breadth of discussion, but by the clarity of direction, inclusiveness of participation, and credibility of follow-through. First, a key outcome should be a globally endorsed, action-oriented framework that moves beyond principles to implementation pathways, clearly outlining how countries, especially emerging economies, can operationalize safe and trustworthy AI within their unique contexts. Second, the Dialogue should produce a commitment to inclusive global participation in AI governance. Africa must not remain a policy taker. Success means creating structured mechanisms, such as advisory groups, regional nodes, or permanent seats, that ensure African voices, institutions, and innovators are consistently represented in global decision-making processes. Third, there should be tangible commitments toward capacity and infrastructure development. This includes funding mechanisms, technology transfer pathways, and partnerships that support local AI ecosystems, particularly in areas like compute infrastructure, data systems, and talent development. Fourth, the Dialogue should catalyze harmonization without homogenization, encouraging interoperable governance approaches while preserving national and regional flexibility. For Africa, this is critical to enable cross-border innovation under frameworks such as the African Continental Free Trade Area (AfCFTA). Hence, success requires a clear accountability and continuity mechanism. The Dialogue should not be a one-off event, but the foundation of an ongoing process with measurable milestones, periodic reviews, and transparent reporting. Ultimately, a successful outcome is one where Africa is not only protected from the risks of AI but is actively empowered to shape, build, and benefit from its future.

• Safe, secure and trustworthy AI
• AI capacity-building
• Interoperability of governance approaches
• Transparency, accountability, and human oversight

My selected priorities reflect both the realities of Africa's emerging AI ecosystem and the practical insights we have gathered through initiatives like the Nigeria Innovation Summit, where government, industry, academia, and startups converge to shape the continent's digital future. 1. Safe, secure, and trustworthy AI is foundational in contexts where digital systems are rapidly scaling without commensurate safeguards. In Nigeria and across Africa, trust will determine adoption. We see an urgent need to embed ethical standards, data protection, and risk management frameworks early, rather than retrofitting them after harm occurs. 2. AI capacity-building is equally critical. The gap is not just technical skills, but also policy literacy, institutional readiness, and public awareness. Through our engagements, we observe that many stakeholders, especially MSMEs and public sector actors, are willing but underprepared. Structured capacity development is essential to ensure inclusive participation and avoid deepening existing inequalities. 3. Interoperability of governance approaches speaks to Africa's fragmented regulatory landscape. Startups and innovators often operate across borders, yet face inconsistent policies. We advocate for harmonized, flexible frameworks that enable cross-border innovation while respecting local contexts. This is key for scaling African solutions globally. 4. Transparency, accountability, and human oversight are necessary to align AI systems with societal values. In our dialogues, stakeholders consistently emphasize the need for explainable systems and clear accountability mechanisms, particularly in high-impact sectors like finance, healthcare, and governance. Together, these priorities reflect a balanced approach: building trust, strengthening capacity, enabling collaboration, and ensuring responsible oversight toward a strong and inclusive AI ecosystem.

Yes. While the listed themes are critical, several cross-cutting and emerging issues require urgent attention, particularly in the context of Nigeria and Africa. First is data sovereignty and equitable data value chains. Much of Africa's data is extracted, processed, and monetized outside the continent, limiting local value creation. Beyond governance, there is a need to ensure that data generated within African contexts contributes to domestic innovation, economic growth, and local AI model development. Second is AI infrastructure inequality. Discussions on capacity-building often overlook the foundational gap in compute power, cloud access, and reliable connectivity. Without deliberate investment in shared digital infrastructure, African innovators risk remaining consumers rather than producers of AI solutions. Third is the informal economy and AI inclusion challenge. A significant portion of Nigeria's economy operates informally, yet most AI systems are designed for formal, data-rich environments. There is an emerging need to design AI that reflects local realities. That is, low-data contexts, vernacular languages, and informal market structures, so that innovation is truly inclusive. Fourth is public-private co-creation as a governance model. Traditional top-down regulation is insufficient for fast-evolving technologies like AI. From our experience convening stakeholders, more adaptive, collaborative governance mechanisms, such as regulatory sandboxes and innovation testbeds, are essential but still underemphasized globally. Lastly, AI for developmental priorities remains underrepresented. In Africa, AI must be intentionally aligned with sectors like agriculture, healthcare, education, and climate change. The question is not just how to govern AI, but how to direct it toward solving high-impact, context-specific challenges. Addressing these issues will ensure that AI governance is not only protective but also transformative and development-oriented for Africa and other developing economies.

In Nigeria and across Africa, governance gaps in AI are already shaping both risks and opportunities across the public and private sectors. A major challenge is the absence of clear, adaptive regulatory frameworks for safe and trustworthy AI. While digital adoption is accelerating, safeguards around data protection, algorithmic bias, and system accountability remain jagged. This creates uncertainty for innovators and increases the risk of misuse, particularly in sensitive sectors like finance, public services, and elections. Another critical gap is limited institutional and human capacity. Many policymakers and regulators are still building foundational understanding of AI, making it difficult to design or enforce effective oversight. At the same time, startups and MSMEs often lack the technical and financial resources to implement responsible AI practices, creating a bumpy playing field. Fragmented governance approaches across African countries further complicate progress. For businesses operating regionally, inconsistent policies increase compliance costs and slow innovation. This fragmentation limits the continent's ability to scale solutions and compete globally. However, these gaps also present significant opportunities. Nigeria and Africa have the advantage of "leapfrogging" legacy systems, allowing us to embed governance, ethics, and interoperability into AI ecosystems from the outset. There is also a growing opportunity to co-create context-specific frameworks that reflect local realities, such as informal economies, multilingual populations, and low-data environments. Additionally, the increasing global focus on responsible AI creates space for strategic partnerships, funding, and knowledge exchange, which can accelerate capacity-building and infrastructure development. In the end, addressing these governance gaps can position Africa not just as a participant, but as a shaper of inclusive and future-ready AI systems, aligned with both global standards and local priorities.

The AI Dialogue can play a catalytic role by shifting international cooperation from fragmented, principle-based conversations to coordinated, and implementation-driven collaboration. First, it can serve as a neutral global convening platform that aligns diverse stakeholders, namely, governments, the private sector, academia, and civil society, around shared priorities while bridging the gap between developed and emerging economies. For Africa, this is critical to ensure that cooperation is not extractive, but mutually beneficial. Second, the Dialogue can drive practical interoperability of governance frameworks. Rather than pushing uniform regulations, it can facilitate mutual recognition models, shared standards, and regulatory mapping tools that allow countries to collaborate without sacrificing sovereignty. This would significantly reduce friction for cross-border innovation and digital trade. Third, it can mobilize collective investment in global public goods for AI, including open datasets, compute infrastructure, safety research, and benchmarking tools. Many African countries cannot independently finance these at scale, but through coordinated international efforts, access can be democratized. Fourth, the Dialogue can institutionalize knowledge exchange and capacity partnerships. This includes peer learning among regulators, technical training programs, and secondment opportunities that embed expertise across regions. Such mechanisms are essential to close the governance and skills gap. Fifth, it can establish accountability and continuity frameworks for global cooperation. By defining measurable targets, periodic reviews, and transparent reporting, the Dialogue can ensure that commitments translate into sustained action. At the end of the day, the AI Dialogue's role is to move cooperation from aspiration to execution—creating a system where countries like Nigeria are not only recipients of global norms, but active contributors to shaping them, while benefiting from shared progress.

The AI Dialogue should deliberately build on, rather than duplicate, existing global and regional efforts that are already shaping AI governance. At the international level, it can connect with the work of UNESCO's Recommendation on the Ethics of Artificial Intelligence, the OECD AI Principles, and the G7 Hiroshima AI Process. These frameworks have established important normative foundations, but the gap remains in translating principles into coordinated, globally inclusive implementation mechanisms. It should also align with emerging technical and safety-focused initiatives such as the Global Partnership on AI (GPAI), the work of the International Telecommunication Union (ITU) on AI standards, and ongoing efforts within ISO/IEC working groups. These platforms are critical for technical harmonization, but often lack strong representation and contextual input from developing economies. Regionally, the Dialogue should interface with African-led efforts such as the African Union's Digital Transformation Strategy and evolving discussions around continental data governance and digital public infrastructure. These are essential entry points for ensuring that Africa's priorities are not peripheral, but integrated into global architecture. The added value of the AI Dialogue lies in its potential to serve as a meta-coordination platform, connecting these fragmented initiatives into a coherent, inclusive ecosystem. It can provide the missing bridge between normative frameworks, technical standards, and implementation realities, particularly for countries with limited institutional capacity. For regions like Africa, its greatest value would be in enabling structured participation, resource alignment, and practical pathways for adoption, rather than leaving countries to independently interpret and implement global principles. In doing so, the Dialogue can help transform AI governance from a set of parallel conversations into a more integrated, equitable, and action-oriented global system.

Different stakeholders can contribute most effectively to the AI Dialogue if participation is structured around clear roles, sustained engagement, and co-creation rather than consultation alone. Governments should provide policy direction and ensure alignment with national and regional priorities, but also actively participate in peer-learning exchanges, where regulators from different regions jointly test approaches such as AI risk classification, sandboxing, and audit frameworks. This is particularly important for emerging economies like Nigeria, where regulatory capacity is still evolving. The private sector, including startups and large technology firms, should contribute through real-world use cases, technical standards input, and responsible deployment practices. Their involvement should go beyond presenting innovations to co-developing governance tools that are practical and scalable across contexts. Academia and research institutions should play a central role in evidence generation, impact assessment, and foresight analysis, helping to anticipate risks and opportunities. Civil society should ensure that ethical considerations, inclusion, and human rights perspectives remain central, particularly for vulnerable populations. From experience convening multi-stakeholder platforms such as the Nigeria Innovation Summit, one key lesson is that impact increases significantly when dialogue is linked to action-oriented outputs rather than being purely discursive. In terms of structure, the AI Dialogue should adopt a multi-layered format: • A high-level global plenary for strategic alignment • Thematic working groups focused on actionable domains (e.g., safety, capacity, infrastructure, governance interoperability, and so on) • Regional hubs to contextualize global discussions, ensuring that African and other regional realities shape outcomes • Innovation labs or sandboxes where policies and tools can be tested in real time The Dialogue should include a continuity mechanism, ensuring outputs feed into measurable commitments, annual reviews, and iterative improvement cycles. This would transform it from a one-off event into a living governance process.

Several critical voices remain underrepresented in global AI governance discussions, particularly from Africa and other emerging economies. This imbalance risks producing frameworks that are globally uniform in aspiration but irregular in relevance and impact. First, informal economy actors are largely absent, despite representing a significant share of livelihoods in Africa. AI systems are increasingly shaping access to credit, markets, and services, yet the lived realities of informal traders, artisans, and micro-entrepreneurs are rarely reflected in governance design. Second, local innovators and early-stage startups are often overshadowed by large technology firms and well-resourced institutions. Yet these actors are closest to context-specific challenges and are experimenting with solutions in agriculture, health, education, and fintech that should inform global standards. Third, policy practitioners and regulators from low- and middle-income countries remain underrepresented in technical standard-setting spaces, despite being responsible for implementation in complex, resource-constrained environments. Fourth, linguistic and cultural communities, particularly those in multilingual African societies, are not adequately represented in AI training datasets or governance debates. This leads to exclusionary systems that fail to serve non-dominant language users. To address these gaps, inclusion must go beyond symbolic participation. The AI Dialogue should institutionalize regional innovation and policy hubs, ensuring sustained African and emerging economy representation in agenda-setting. It should also support funded participation mechanisms, enabling smaller states and grassroots actors to contribute meaningfully without financial barriers. Also, structured community consultations and sector-specific listening sessions, particularly in informal markets and rural areas, should inform global discussions. Finally, integrating outputs from platforms like the Nigeria Innovation Summit into global dialogues can help bridge the gap between grassroots innovation ecosystems and international governance frameworks. True inclusion requires shifting from participation to co-creation of AI governance itself, ensuring that global rules reflect the diversity of real-world contexts in which AI is deployed.

To ensure the AI Dialogue moves beyond traditional conference-style exchanges, it should adopt engagement formats that prioritize co-creation, real-world problem-solving, and sustained interaction across stakeholders and regions. One effective format is Policy and Innovation Sandboxes, where regulators, startups, and researchers jointly test AI governance approaches in controlled, real-world environments. This allows participants to evaluate the practical implications of policies, such as bias audits, transparency requirements, or data governance rules, before they are formally adopted. Another impactful approach is Challenge-Based Dialogues, where global and regional teams work on defined problems such as AI in agriculture, healthcare access, or misinformation. This mirrors innovation models used in platforms like the Nigeria Innovation Summit, where multi-stakeholder teams co-develop solutions rather than simply exchange views. The Dialogue should also include Regional Co-Creation Hubs, particularly in Africa, Latin America, and South Asia. These hubs would ensure that local contexts actively shape global outcomes, rather than being retrofitted into them. Outputs from these hubs should feed directly into global plenaries. A further innovation is Living Governance Labs, where policymakers, technologists, and civil society iteratively refine governance frameworks over time. Unlike one-off consultations, these labs would operate continuously, allowing for adaptive governance aligned with rapidly evolving AI systems. Furthermore, Narrative and Scenario Labs can be used to explore future risks and opportunities through storytelling, simulations, and foresight exercises. This helps stakeholders collectively understand complex trade-offs in AI development. To conclude, the Dialogue should incorporate digital participation platforms with asynchronous engagement tools, enabling broader inclusion of actors who cannot attend physically, particularly from underrepresented regions and informal sectors. Together, these formats would transform the AI Dialogue into a dynamic ecosystem of experimentation, learning, and shared accountability, rather than a static policy forum.

Several practical policy approaches and platforms already demonstrate how AI governance can move from principle to implementation, and these can be scaled or adapted globally. First, AI regulatory sandboxes, such as those used in the UK and Singapore, provide controlled environments where innovators and regulators test AI systems under real-world conditions. This approach reduces uncertainty while enabling evidence-based regulation. For countries like Nigeria, sandboxes are particularly valuable for fintech, healthtech, and digital identity systems where innovation is rapid but regulatory clarity is still evolving. Second, risk-based AI governance frameworks, such as the EU AI Act approach, offer a structured way to classify AI systems based on levels of risk and apply proportionate safeguards. This helps avoid over-regulation of low-risk innovation while ensuring strict oversight for high-impact applications. Third, digital public infrastructure (DPI) models, such as India's Aadhaar and Unified Payments Interface (UPI), demonstrate how interoperable, state-enabled digital systems can accelerate inclusion while enabling governance at scale. For Africa, similar DPI approaches in digital identity, payments, and data exchange can support both innovation and accountability. Fourth, open standards and interoperability frameworks, promoted by organizations like the ITU and ISO/IEC, are essential for ensuring that AI systems can work across borders and platforms. This is particularly relevant for Africa under the African Continental Free Trade Area (AfCFTA), where fragmented systems remain a challenge. Lastly, multi-stakeholder platforms like the Nigeria Innovation Summit demonstrate the value of co-creation ecosystems, where policymakers, innovators, academia, and industry jointly identify challenges and shape solutions in real time. Collectively, these approaches highlight that effective AI governance is not only about regulation, but about building adaptive systems that integrate experimentation, infrastructure, inclusion, and collaboration.