Technology Learning and Building Solutions (takenoLAB)

For the first Global Dialogue on AI Governance to be truly successful, it must produce outcomes that reaches those currently left behind in the AI revolution. First, the Dialogue should result in clear policy commitments that give all nations (rich or poor, large or small) an equal and genuine voice in how AI is governed globally. Decision-making structures must reflect the full diversity of the world, not just the interests of a powerful few. Second, success would mean launching dedicated support programs for smaller organizations and developing countries that lack access to the computing power needed to build AI. Wealthier governments, international bodies, and private companies should commit to funding or sharing these resources, so that more countries can develop and use AI on their own terms. Third, the Dialogue should start a global movement to make AI openly available to everyone. This means committing to ensure that key AI tools, data, and knowledge are treated as shared global resources, free to access, not locked away by large corporations or wealthy nations. The goal should be to move from a world where AI power is concentrated in a few hands, to one where its opportunities are shared by all.

• AI capacity-building
• Open-source software, open data and open AI models
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Transparency, accountability, and human oversight

My selection of priorities is guided by a simple belief that AI should work for everyone, not just those who already have power and resources. AI Capacity-Building is my top priority because access to AI is still deeply unequal. Many countries and organizations want to participate in the AI economy but lack the basic tools, computing power, and skills to do so. Without targeted support to close this gap, global AI governance will remain a conversation dominated by a few wealthy nations and large corporations. Open-source software, open data and open AI models give smaller nations and organizations a real chance to build, adapt, and benefit from AI on their own terms. When AI tools and knowledge are locked behind paywalls or proprietary systems, they reinforce existing inequalities. Social, Economic, Ethical, Cultural and Linguistic Implications of AI matter because the impact of AI is not the same across all communities. We can already see how languages and communities are underrepresented in large AI models, largely due to limited data. Now, as AI becomes a trusted source of information and is incorporated into many day-to-day activities, it becomes obvious that there is a need to push for fair and equal representation of all nations and languages. Transparency, Accountability, and Human Oversight are essential because developing nations and smaller organizations are often on the receiving end of AI systems they did not build and cannot inspect. Accountability and human control will help close the gap between those who build AI and those who use it.

Yes, there are important issues that the listed themes do not fully address. Data Ownership and Sovereignty is a top concern because every day, developing countries are generating large amounts of data with little or no control over how that data is collected, used, or profited from. As we now know, large AI companies use this data to train their models without fair compensation or consent. Global AI governance must directly address data ownership and ensure that everyone can benefit from collected data, regardless of where a nation or organization stands in the AI ecosystem. Concentration of Power is another critical issue. While governments play a role in AI today, the real decision-makers in this industry are a small number of powerful corporations. This concentration of power is dangerous because it means that governance frameworks for this massive industry risk being shaped by corporate interests rather than the public good. Workforce Displacement is an unavoidable reality that we must plan for now. Whether we like it or not, AI will continue to disrupt jobs and industries. It is important to think ahead and put in place safety nets that help people and communities manage this change, a change we are already living through, and one that will only grow as AI continues to expand.

Capacity-building comes first, as the gap here is very wide. Without global commitments to share computing resources, funding, and technical training, developing countries will continue to fall further behind. The key opportunity would be targeted investment in local capacity to unlock talent and innovation that large corporations often overlook. Open-Source and Open Data address another pressing challenge. The majority of developing nations spend large portions of their budgets on expensive licensed AI tools that do not represent them and offer no sense of ownership to these nations and organizations. Open-source and open data provide the entry point for everyone to participate in training models that represent their nations using their own languages and on their own terms. Social, Cultural, and Linguistic Implications are already visible on the ground. Even though AI tools have been deployed in environments such as education, healthcare, and public services, they frequently fail to support communities that speak minority languages. This is the opportunity to shift ownership away from large corporations and toward smaller communities, empowering them to build and train models that truly understand and serve their languages and needs.

The AI Dialogue should help establish shared rules that everyone agrees on. This will give all nations a clear starting point when developing their own AI governance policies. Second, it should amplify the voices of developing countries and nations that are eager to build their own AI tools but are struggling with computing power and access to resources. Third, it should highlight and connect the efforts of different countries, regions, and organizations that are already working toward the common goal of making AI accessible to everyone.

UN Secretary-General's High-Level Advisory Body on AI (HLAB-AI) Global Partnership on AI (GPAI) UNESCO Recommendation on the Ethics of Artificial Intelligence

AI Builders and Solution Developers, startups, local tech organizations, and innovators building real AI tools on the ground should have a dedicated space to share what is working, what barriers they face, and what support they need to scale their solutions responsibly. Governments should come with clear commitments, especially developing nations who must be supported with translation and technical assistance to participate as equals.

Minority language communities are rarely represented in either AI development or governance discussions. Indigenous communities hold unique knowledge systems and face specific risks from AI including misrepresentation and data extraction without consent. Small and local AI builders, developers and innovators working in resource-limited environments rarely get to influence global governance conversations despite building solutions closest to real community needs. They should have a formal voice in how rules are made.

Instead of listening to speeches, participants should work together in small, mixed groups to tackle real AI governance challenges. Local and emerging AI builders from developing nations should be given a platform to demonstrate their solutions, share their barriers, and connect with potential partners and funders.

Mozilla Common Voice This open platform crowdsources voice data in minority and underrepresented languages. It shows how open data collection can be done inclusively, giving communities ownership over how their languages are represented in AI systems. Takenolab, AI Training for Refugees in Malawi Takenolab is delivering practical AI training to refugees in Malawi, proving that meaningful AI capacity-building can reach even the most marginalized and resource-limited communities. This is exactly the kind of grassroots initiative that global AI governance frameworks must support, protect, and scale.