Ns Law Chambers

The successful first Global Dialogue on AI Governance, in my view, should feel less like a meeting and more like a starting point for real change. It should mean that countries from the Global South are not just present in the room, but actually influencing the direction of the rules being shaped. It would also be important to come out with a few simple shared principles, things like fairness, transparency, and respect for human rights,but without forcing every country into the same system. Different places will need different ways of applying them. Another real outcome would be support for countries that are still catching up. Many institutions simply don't yet have the tools or expertise to regulate AI properly, so capacity building has to be part of the result, not an afterthought. I also think data needs to be part of the conversation in a practical way,especially how it moves across borders and how countries can still benefit from data generated within them. And finally, there should be a clear sense of "what happens next." If nothing follows after the Dialogue, then even good discussions will not mean much. In the end, success is really about whether this process leads to something people can actually see and use in real life.thank you

• Safe, secure and trustworthy AI
• Transparency, accountability, and human oversight
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• AI capacity-building

1. Safe, secure, and trustworthy AI This is important because AI is already being used in serious areas like law, finance, and public services. If it is not properly checked, it can easily lead to mistakes or unfair outcomes, so people need to be able to trust it. 2. AI capacity-building In many places, like Tanzania, we are still learning how AI works in practice. Without building skills and understanding, it is hard for institutions to regulate properly, let alone keep up with it. 3.Transparency, accountability, and human oversight People should not be left guessing how AI decisions are made. There needs to be clarity, responsibility, and real human control, especially when decisions affect people's lives. 4.Social, economic, ethical, cultural, linguistic, and technical implications of AI AI is changing everyday life in many ways-how people work, communicate, and access opportunities. But many systems do not reflect local languages or realities, so AI must work for all societies, not just a few.

Yes, there are a few important issues that are not fully captured in the listed themes. One is digital inequality, both between countries and within them. In reality, some places have strong access to data, infrastructure, and skills, while others are still far behind. This gap affects who benefits from AI and who is left out. Another issue is data ownership and value. A lot of the data used to train AI comes from people in developing countries, but the real value is often created and kept elsewhere. This raises fair questions about who benefits from that data. There is also the environmental impact of AI, especially the high energy use from large systems and data centres. This is not talked about enough, but it is becoming more important as AI keeps growing. We also see more informal use of AI tools, where systems are being used in workplaces and public services without clear rules or oversight. This can create confusion about responsibility when things go wrong. Finally, there is the issue of public trust and understanding. Many people are affected by AI decisions but don't really understand how those decisions are made. If people don't trust the system, then even good laws may not be enough. Overall, these issues show that AI governance is not only about regulation, but also about fairness, awareness, and making sure no one is left behind.

In Tanzania, you can already feel the gap between how fast technology is moving and how prepared our systems are to manage it. Take mobile loans for example. Many people in Dar es Salaam or upcountry use loan apps through mobile money. Someone can apply and get a decision in seconds,but sometimes they are rejected without any reason. A small shop owner in Kariakoo or a boda boda rider in Mwanza may not understand why, especially when there is no human to explain or review the decision. In the legal field, we are also not fully digitized yet. Many court records and case laws are still not easily available in digital form. So if AI tools are used for legal research, they may miss important local cases or rely on foreign information that doesn't really fit our laws. In practice, that can lead to wrong advice if not carefully checked. Even in government services, you see a mix,some offices are using digital systems while others are still fully manual. This creates confusion and inconsistency for citizens who just want simple services done quickly and fairly. At the same time, there are clear opportunities. Tanzania has a strong mobile money culture, and people are already comfortable using digital tools. If AI is introduced in a responsible way, it could really help farmers, small businesses, and even hospitals work more efficiently. But overall, the reality is this: people are already being affected by AI-like systems in their daily lives, even though the rules, understanding, and oversight are still catching up.

The AI Dialogue can really help make global discussions on AI more practical and less theoretical. It can create a space where countries actually talk openly and build trust under the United Nations instead of working separately or in isolation. It can also help countries agree on a few simple shared principles like safety, fairness, transparency and respect for human rights while still allowing each country to apply them in a way that fits its own reality. Another important role is supporting countries that are still building capacity. For example countries like Tanzania may need more training, tools and technical support to properly understand and regulate AI systems. It can also help countries deal with cross border issues like data flows and AI systems that operate across multiple jurisdictions because no single country can handle these challenges alone. Most importantly it should make sure that countries from the Global South are not just in the room but actually influencing the decisions being made. In simple terms the Dialogue can help make AI governance more connected, more fair and more practical for everyone involved.

The AI Dialogue should not start from scratch. It should build on what is already there and connect the dots. At the global level, it can link with the work already being done under the United Nations, especially the Global Digital Compact on digital cooperation and AI. It should also connect with other efforts like UNESCO's work on AI ethics and regional digital initiatives in Africa. In countries like Tanzania, we already have frameworks such as the Personal Data Protection Act, 2022 that are starting to shape how data is handled. The real value of the Dialogue is bringing all these efforts together in a way that actually makes sense in practice. Right now, a lot of work is happening in different places without strong coordination. It can also help countries learn from each other in a more practical way, especially where some are more advanced in AI and others are still building systems and skills. Most importantly, it should make sure countries from the Global South are not just following rules made elsewhere, but are also part of shaping them.

Different stakeholders should take part in a very practical way, sharing real experiences instead of only formal statements. Governments can talk about the real challenges they are facing in regulating AI and digital systems, including issues like data protection under the Personal Data Protection Act, 2022 and how hard it is to keep up with fast-moving technology. Civil society and legal professionals can bring in the human side of things, especially how AI is affecting people's rights in everyday life like privacy, fairness, and access to justice. The private sector should be more open about how their AI systems actually work, especially when it comes to transparency, bias, and accountability when things go wrong. Researchers and experts can help by breaking down complex issues and turning them into simple, practical ideas that policymakers can use. As for the Dialogue itself, it should feel less formal and more interactive, with smaller group discussions, regional conversations, and real examples from countries. There should also be follow-up so that ideas don't just stay in meetings but lead to real action. Most importantly, everyone should be properly included, especially countries from the Global South, so they are not just listening but actively shaping the direction of AI governance.

One group that is really missing from global AI discussions is people in rural areas. In places like rural Tanzania for example Dodoma, Kigoma or other villages people are already feeling the impact of AI without even calling it that. It could be through mobile money loans, farming advice tools, or digital services. But the truth is, they are almost never part of the conversations where these systems are designed or decided. A farmer might receive a decision from a mobile loan app or an agriculture platform, but there is no one to explain how that decision was made or whether it really fits their situation on the ground. The problem is that most of these global discussions happen in formal, technical spaces that are far away from everyday life. That makes it hard for rural voices to be heard. To fix this, the AI Dialogue needs to go beyond conference rooms. It should involve local leaders, community groups, and simple, accessible ways for people in rural areas to share their experiences. In simple terms, if AI is really going to work for everyone, it has to include the people who are usually the last to be heard but often the most affected.

At the global level, the Global Digital Compact under the United Nations is a key starting point because it promotes shared principles like human rights, safety, and inclusion in the digital space. In Tanzania, the Personal Data Protection Act, 2022 is an important step in regulating how personal data is collected and used, especially as AI systems become more data-driven. Practical approaches like regulatory sandboxes are also useful, where new digital or AI products are tested in a controlled environment before full release. In Tanzania, similar informal approaches already exist in fintech and mobile money regulation through bodies like the Bank of Tanzania and the Capital Markets and Securities Authority. AI ethics guidelines help ensure systems are fair, transparent, and respect human rights, while open-source tools and data platforms improve transparency and collaboration. Finally, multi-stakeholder engagement is important because AI governance requires input from government, private sector, civil society, and experts. Overall, these examples show that effective AI governance is a combination of laws, practical tools, and cooperation rather than a single system.