“UN Hands-On AI Day” - Leading with Purpose in the Age of AI

Several weeks ago, I had the privilege of speaking at the first UN Hands-On AI Day - an event that brought the United Nations community together to explore how Generative AI is transforming the UN work. From inspiration to action, this event was promoted as the one marking a “pivotal moment in the UN’s journey towards harnessing emerging technologies”. Under-Secretary-General Amandeep Gill, in his welcome address, highlighted the larger shift towards productivity and curiosity at the UN, emphasizing how AI tools can enhance the UN's mandate.

“Today’s event is part of a larger shift towards productivity and curiosity at the UN to imagine how these tools can help our mandate,” he stated.

Assistant-Secretary-General and Chief Information Technology Officer Bernardo Mariano Junior followed with opening remarks that set the tone for the day:

“My hope is that you walk away with new ideas, practical next steps—and a sense of excitement about what’s possible.”

These words reflected the excitement and potential of this gathering, which combined learning, demonstrations, and future-focused dialogue.

After the panel discussion, different teams across the UN Secretariat showcased how they are already utilizing Generative AI in their work. The interactive exhibition featured eight kiosks, each designed to engage participants in hands-on demonstrations and discussions about the potential of AI. These prototypes, at various stages of development, highlighted the diverse applications of GenAI within the UN system and provided a glimpse into how AI can accelerate the UN’s mission. “Empowering Global Engagements on the SDGs”, “AI Governance and Enablement”, “UN Docs Assistant”, “UN Internal Justice Bot”, to name the few, each kiosk offered valuable insights into how AI is transforming work in the UN. The event also provided opportunities for attendees to enroll in UN 2.0 communities, fostering further collaboration.

With over 2,000 people registered online and a packed in-person turnout, the event demonstrated strong interest, not just in learning about AI applications at the UN, but also in gaining hands-on experience. It was a pleasure to exchange ideas on how we can continue advancing AI for the UN and for the global good. In this article, I will share insights from my visit along with recommendations on how the UN can leverage Generative AI (GenAI) to advance its work, and especially the work around the Sustainable Development Goals (SDGs).

From AI Research to Real-World Impact

For the past ten years, the focus was on building the science of AI, developing the core technology. Now, we’re entering a new phase: the next decade will be all about applying AI to real-world problems. This transformation is evident in recent developments - Nobel Prizes in Physics and Chemistry recognized AI-related breakthroughs in deep learning and protein folding. Governments are reflecting this shift too. The AI Action Summit in France, held in February, also marked a shift, with an emphasis on how AI can be used for practical, industry solutions, shifting the emphasis from “safety” to “action”.

Even before this GenAI wave, we saw strong applications in the field I am working in, and the UN is concentrating its efforts, peacebuilding, human rights, and humanitarian action - like predicting he probability of conflict or violence happening, supporting humanitarian response, and enabling anticipatory action. These data-driven efforts previously needed deep technical skills, which is why we started building capacity and creating new jobs, including in the UN that didn’t exist before. But with generative AI, things are changing substantially. Yes, technical skills still matter - but now, everyone can play a role. In the next ten years, we’ll all become part of putting AI to work.

Why the UN can’t sit on the sidelines, is one of the questions we tackled in the discussion. It’s not just that the UN shouldn’t stay out of AI - it won’t be able to. With shrinking resources, budget cuts, staff shortages, and fast-changing global challenges, the UN will have to look for allies. AI, especially GenAI, can be one of those allies if used thoughtfully and strategically. But let’s stay grounded – GenAI is one of many waves of AI we’ll see in the future. The real challenge is shifting our mindset and building flexible systems that can adapt, not just chasing the latest shiny thing.

GenAI and the SDG Gap: A Promising Tool, But No Silver Bullet

As we enter a critical time for achieving the Sustainable Development Goals (SDGs), generative AI (GenAI) offers a promising - though not definitive - opportunity to accelerate progress. Its ability to generate, translate, recognize, organize, and synthesize information can help expand access to essential services, particularly in underserved areas. GenAI can localize educational materials, assist with real-time translation, and provide basic legal aid through chatbots, making it a valuable tool where human expertise is limited.

Beyond service delivery, GenAI can enhance decision-making by helping close data gaps and enabling faster, more informed responses. It can simulate policy scenarios, interpret complex datasets, and generate insights that support more agile and effective governance. This is especially relevant for UN agencies and governments working to respond to evolving crises or implement evidence-based programs at scale.

GenAI also holds potential for advancing inclusive governance under goals like SDG 16, which focuses on peace, justice, and strong institutions. It can provide multilingual legal aid tools, analyze public procurement data to flag corruption risks, and support early warning systems for conflict using insights from local reports and social media. While not a silver bullet, GenAI, if used responsibly, can help scale innovative, locally relevant solutions to some of the world’s most persistent development challenges.

Recognizing the growing importance of generative AI and aiming to prepare its staff for its responsible use, the United Nations Office of Information and Communications Technology (OICT) has published a Generative AI Primer. The primer offers a concise introduction to generative AI, outlining both its benefits, such as efficiency, personalization, creativity, and innovation, and the broader implications for the UN’s work. It is part of OICT’s wider effort to build an AI governance framework and provide guidance on how generative AI can be used within the UN, with a focus on enhancing information discovery, automating knowledge management, and expanding access to education and training.

The Paradox of AI applications for SDGs - Apart from risks of misinformation, misuse, or unintentional harms such as hallucinations of these tools, one major risk is deepening inequality. AI might help achieve SDGs in some countries but leave others behind. Let’s not forget: 2.6 billion people still don’t have internet access. That’s a third of the world left out of the AI revolution. If AI is developed without considering low-income or Global South contexts, we risk deepening the digital divide.

Ask the SDG experts, not the private sector, how AI can be used to advance the SDGs

I advise caution when turning to the private sector for answers. One of our first mistakes in exploring AI as a solution for the SDGs, especially under such tight time constraints, is that we often ask AI experts how AI can help achieve SDGs. The problem with that approach is that many of these experts know little, if anything, about the world’s most pressing development challenges, though there are, of course, exceptions.

At the end of last year, I listened to the UN Security Council ministerial briefing on AI and its role in “Maintaining International Peace and Security.” Two of the field’s most respected figures, often referred to as the godmother and godfather of AI, spoke at the event. It was striking, and frankly painful, to witness how little they understood about global peace and conflict dynamics, or the harm being inflicted on vulnerable populations, some of which is directly facilitated by AI systems.

When Sam Altman was asked recently during his interview at the TED 2025, what kind of future he wants for his child, he replied, “I want a future in which computers will be smarter than my kid.” With respect to his work and achievements, I imagine anyone working on the Sustainable Development Goals would have a very different answer - one focused not on machines, but on people. A world without war or hunger. A world where no child dies in conflict, where every child can go to school, grow up healthy, and thrive on a heathy planet, in a just and peaceful society.

This is why we need to reverse the conversation and start by asking those working on the SDGs, especially those with lived experience from countries that are falling behind, what their needs and challenges are. Only then can we begin to identify where AI might genuinely help. This also means demystifying AI for these communities so they can assess for themselves where meaningful applications exist. At the same time, we must build trust and transparency to ensure that AI tools protect, rather than exploit, vulnerable populations.

What the UN Can Learn from Other Sectors

I really enjoyed tackling this question, as my work involves engaging with the entire ecosystem, including civil society, academia, governments, the private sector, and the UN. Academic institutions are especially strong at combining AI with domain expertise and often operate transparently, sharing their research openly with the broader community. The UN could take inspiration from this model by fostering interdisciplinary and transparent partnerships, not only within its agencies but also across sectors. While I was inspired by some of the AI initiatives showcased at UN Headquarters, that knowledge is too often kept within internal circles.

As our hosts noted in their opening remarks, governments are likely to be the next frontier for impactful GenAI applications, already testing tools in real-world policy settings through innovation labs and regulatory sandboxes. The UN could adopt a similar approach to pilot AI in peacekeeping, development, or humanitarian response before scaling solutions globally.

However, when it comes to the private sector, we’ve already seen the consequences of the "move fast and break things" mentality, especially in fragile contexts where things do, in fact, break, if not developed and implemented with those specific contexts in mind. The misuse and unintended consequences of social media in the vulnerable settings should serve as a clear warning. With AI, we now have the opportunity to take a different path, one grounded in foresight, responsibility, and equity. The UN has a critical role to play in shaping that path and ensuring AI serves the common good.

Far from being a newcomer, the UN has been at the forefront of this work

Yes, the UN can learn from how others innovate. But those same sectors also have a lot to learn from the UN, especially on ethics, equity, global governance, and long-term thinking. The UN offers something unique that the private sector, academia, and governments cannot: a global mandate rooted in human rights, peace, and inclusion. The UN has the power and responsibility to ensure that AI is developed in a way that aligns with these values.

For the last ten years, I’ve been working with different UN agencies across the Humanitarian-Development-Peace nexus, and I can share my experiences that the UN is often a leader and not a newcomer in emerging technologies, including AI. For several years, I’ve been organizing “Data for Peace” conferences in NYC and online, where UN has been showing a leadership in early warnings in the age of compound risks, peacekeeping early waring and response tracking, the power of messaging apps in humanitarian settings (long before the GenAI advancements), to only mention a few which you can still check online on NYU CIC Youtube Chanel. And not to forget two major hubs, still providing valuable examples to other sectors of how data can be collected and used responsibly, while advancing data-driven work, modeling and predictions in both humanitarian and development fields: The UN OCHA Centre for Humanitarian Data – Humanitarian Data Exchange and Peace and Security Data Hub, both serving as models when we were envisioning the future of the potential Prevention and peacebuilding Data Hub.

UN Secretary-General António Guterres has a huge role in advancing this agenda, spearheading several initiatives to address the opportunities and challenges posed by emerging technologies, particularly AI. Notable initiatives include: UNSGs Strategy on new Technologies (2018), Strategy for the Digital Transformation of UN Peacekeeping (2021), and more recently Global Digital Compact and the High-Level Advisory Body on AI (2023). And of course, not to forget the UN 2.0, a modernization agenda launched by UNSG to equip the UN with the tools and mindsets needed to address 21st-century challenges, focusing on data, digital transformation, innovation, foresight, and behavioral science. It aims to make the UN more agile, forward-looking, and effective in delivering on its mandates, including the SDGs.

The UN Hands-On AI Day was more than just a showcase - it was a signal of intent. As generative AI reshapes the global landscape, the UN has both an opportunity and a responsibility to lead with purpose. By grounding innovation in lived experience, equity, and global collaboration, the UN can ensure that AI serves not just progress, but people. The moment calls for bold leadership, thoughtful experimentation, and a commitment to using AI as a force for inclusion, peace, and sustainable development.

Welcome to Is AI Coming in Peace? - a personal newsletter where I explore the complex and evolving relationship between artificial intelligence and peace. From its potential to amplify peacebuilding, humanitarian efforts and prevent violent conflicts to the grave risks it poses when misused, AI is reshaping the way we address the world's most pressing challenges.

As someone deeply committed to building peace, I’ll share reflections on AI’s role in global issues, lessons from projects I am working on, events I’ve attended, and readings that attracted my attention. It’s a mix of personal insights and professional explorations into how we can ensure this powerful technology contributes to peace rather than war.