Our stories. Find out more about our work towards the Sustainable Development Goals.

I was recently privileged to participate in three landmark global gatherings on artificial intelligence (AI) in Geneva: the AI Governance Global Dialogue, the AI for Good Global Summit, and the World Summit on the Information Society (WSIS)+20 High-Level Event. Together, these forums brought governments, technology leaders, researchers, civil society and the United Nations together to grapple with one of the defining questions of our generation: how do we ensure that artificial intelligence and other emerging technologies serve humanity?A Moment of Promise and CautionI returned from Geneva with feelings of excitement and concern.Excitement at AI’s extraordinary potential to improve lives. Every discussion and exhibition revealed breakthroughs that, only a few years ago, would have seemed unimaginable. Capabilities are expanding not over decades, but months, transforming healthcare and education, increasing agricultural productivity, strengthening disaster preparedness, improving public services and accelerating scientific discovery. Indeed, the World Bank’s 2026 World Development Report focuses on ‘The Promise of Artificial Intelligence’ and argues that AI can help developing nations achieve a century of progress in just a decade provided developing countries make the right investments in infrastructure, electricity and digital skills.But I also returned deeply concerned. Not because AI itself is something to fear, but because technological progress may be advancing faster than our systems to safely harness and govern it. The AI Governance ImperativeAI governance is not simply a technological debate; it is a matter of safety, human agency and, ultimately, human rights.  The real question is whether our institutions, laws and systems of international cooperation can evolve quickly enough to ensure that AI remains safe, inclusive, beneficial to everyone and environmentally sustainable.In thinking about AI safety, I find everyday analogies useful. We board a Kombi, put our children on a bus or drive onto a busy road with confidence because we rely on agreed and globally recognizable rules - the traffic code. While human error and mechanical faults can and do cause costly failures, these rules provide an agreed basis for accountability and improving safety. The same applies to aviation: we board aircraft because we trust the rules governing the industry globally. Without the layers of safety and coordinated regulation governing the industry, flying would be a death wish.Given the tremendous impact of AI and other emerging technologies, it should not be left unregulated or solely to voluntary self-regulation by the companies developing them.Concerns about human oversight, safety, privacy, accountability, cultural and linguistic biases or exclusion, and data inaccuracies, are real and urgent. While a single global regulator or universal model may still be some distance away, interoperability among national, regional and international frameworks is an important starting point. These frameworks must work together, uphold shared principles and respect different legal systems and national priorities. Ultimately, however, we must work towards a global framework. The urgency is clear: just days after the Geneva conferences, at least two leading AI companies reported experiencing troubling failures, including models circumventing instructions, breaking away and hacking into other systems, including ‘hijacking’ a website, all the while unnoticed by their handlers until week and months later. A few years ago, this would have sounded like science fiction. Now it is real.While many countries remain cautious about multilateral governance and favour greater national control, AI presents challenges no country can address alone. From cybersecurity and misinformation to cross-border data flows, the risks – and opportunities – extend beyond borders. Multilateral cooperation is therefore not optional; it is indispensable.There is also the critical environmental dimension. AI depends on computing power, requiring significant energy, water and infrastructure. The recently released United Nations University report titled "Environmental Cost of Artificial Intelligence: Carbon, Water, and Land Footprints," reveals the sheer amount of electricity and water needed to drive AI, as well as the resultant carbon footprint.  The report prompted UN Secretary-General António Guterres to launch the AI Environmental Transparency Initiative which seeks to have AI companies publicly disclose their full carbon, water, and land footprints, while committing to power all data centers with 100% renewable energy by 2030.For countries such as Eswatini, where large-scale computing capacity may be limited, we must consider not only how we can harness the power of AI, but also remain fully engaged on global platforms about the safe, secure and sustainable deployment of AI.  Inclusivity: Africa Must Help Shape the FutureDigitalization is one of six investment pathways identified to accelerate the SDGs and is particularly expected to help bridge development gaps by harnessing technology to widen inclusion. Yet current AI trends challenge this assumption. As the UN Secretary-General reminded us at the AI Governance Global Dialogue: “The computing power, the data and the talent behind the most advanced systems are concentrated in a handful of companies, and in a handful of countries. Most nations, including many developing countries, have had no say in decisions that will shape their futures.”  Indeed, AI computing capacity is highly concentrated, with over 90% located in only two countries.  The World Bank World Report cited above proposes an optimistic path where countries can adapt low-cost AI tools to local contexts rather than focusing on setting up expensive models of their own. Such local adaptation must go together with the positioning of indigenous languages and knowledge and the inclusion of cultural diversity that are currently absent from training data.Furthermore, for Africa, while adaptation may be a starting point, efforts must continue to move beyond being viewed as a market, to becoming a partner in shaping the technological future that it wishes to adapt.  Investment in digital infrastructure, connectivity, research and skills, alongside stronger regulatory institutions will be key in ensuring safe adaptation and narrowing the digital divide.A Governance Response for EswatiniThe emerging concerns point to the importance of strengthening governance at the national level. For Eswatini, this journey is already underway: the country is developing a National AI Strategy, aligned with the African Union’s vision and Eswatini’s domestic development agenda. The country is also advancing a combined Fourth and Fifth Industrial Revolution Strategy, aimed at enabling Eswatini to leapfrog developmental gaps and position its economy for the future.At the same time, Eswatini is strengthening its legal and institutional environment to support the responsible adoption of AI. Existing instruments, including the Data Protection Act, the Computer Crime and Cybercrime Act and the Electronic Transactions Act, provide an important baseline, while work towards dedicated AI governance provisions will help address the specific opportunities and risks presented by increasingly powerful technologies.The objective should not be to slow innovation, but to create the conditions for responsible innovation to thrive. As Prime Minister Russell Dlamini emphasized at the AI Governance Global Dialogue, effective AI governance must strike a careful balance: regulation must protect without stifling innovation. The goal is to establish “smart, forward-looking regulatory frameworks” that encourage innovation while safeguarding citizens.Such a framework should help ensure that AI is directed towards Eswatini’s development priorities. Whether in healthcare, education, agriculture, public administration or employment, the objective should be to use technology to improve lives while ensuring meaningful human oversight where decisions affect people’s rights and opportunities.Importantly, this journey cannot be undertaken by Government alone. The private sector, academia, civil society, young people and development partners all have a role to play in shaping Eswatini’s approach to AI. And as AI governance continues to evolve globally, the country should ensure that its national framework is forward-looking and remains connected to regional and international discussions.Human Rights by DesignThe United Nations has consistently advocated for AI that is safe, human-centred, rights-based and inclusive. Through the Global Digital Compact, Member States have committed to ensuring that digital technologies advance sustainable development while protecting human dignity, equality and fundamental freedoms.Turning these principles into reality requires governance that is as innovative as the technology itself.Human rights must remain at the centre of AI governance. Transparency, accountability and meaningful human oversight cannot become optional features added after technologies are deployed. They must be embedded by design.Whether AI is used in healthcare, education, employment, justice or public administration, people must understand how decisions affecting their lives are made and have mechanisms for accountability when harm occurs.There is also a profound ethical responsibility. As the saying goes, "He who builds a bridge has the moral responsibility to ensure that it is safe." The same principle applies to artificial intelligence. Those who design increasingly powerful AI systems bear a corresponding responsibility to ensure they are safe, transparent, secure and aligned with the public good.A Shared FutureBack to the Geneva conferences, what was particularly heartening for me was the resounding affirmation of the United Nations’ role as a convenor. The AI Global Governance Dialogue was a direct outcome of the Global Digital Compact, adopted by Member States alongside the Pact for the Future in 2024. Over 170 member states showed up in Geneva, as did civil society, researchers and technology company leaders. While they did not all see the issue from the same perspective, they engaged constructively on the need for stronger AI governance. The dialogue is on and the United Nations is playing an increasingly important role in convening this global conversation. Through the leadership of the International Telecommunication Union (ITU) and initiatives such as the AI for Good Global Summit, governments, industry, academia and civil society are working together to shape governance approaches that encourage innovation while safeguarding human rights and advancing sustainable development.Artificial intelligence and other emerging technologies are undoubtedly shaping the future. If governed wisely, AI can accelerate progress towards the Sustainable Development Goals, expand opportunity and improve lives. If left ungoverned, it risks overtatking human agency, widening inequality, reinforcing exclusion and concentrating power even further. Its future should not be determined solely by those with the greatest computing power. It should be shaped collectively by humanity.  Our responsibility is to ensure that the future being shaped is equitable, inclusive, transparent and worthy of the generations that will inherit it.If we do not act together, we may look back on this period as the moment when we lost the plot. We must act now.