Social Sciences for AI

About the Event Nordic AI companies are building products for global markets. Their teams overwhelmingly draw on English-language data, Western benchmarks, and a narrow talent pool. A health AI trained on European clinical data produces unreliable results in Nairobi. A conversational AI built on English-language patterns misreads turn-taking norms in Mandarin or Swahili. These failures are predictable. They are also expensive. They keep happening because AI product teams lack the contextual knowledge that would catch them before deployment. Social scientists who study the Global South can see these breakdowns coming. They understand infrastructure gaps, user behaviour, regulatory landscapes, and local market logic that no training dataset contains. For any company scaling AI into high-growth regions, this knowledge has direct commercial value. This panel asks: what does Global South knowledge offer AI companies, what makes it hard to integrate, and what would it take to do it well? Oslo Innovation Week 2026 | Side ProgrammeDate 20.10.2026 15:00-16:00 | Oslo, Norway What We Will Discuss Why Global South knowledge makes AI products betterAI failures in new markets are context problems, not technical problems. Companies that build without local knowledge lose time, money, and trust. We look at concrete cases where contextual expertise changed outcomes, from product localisation to dataset design. What social science brings that engineering does notMultilingual fluency. Fieldwork experience in the markets where AI adoption is growing fastest. Firsthand knowledge of how regulations, infrastructure, and user behaviour vary across regions. This is applied intelligence that gives companies an information advantage their competitors lack. The hard parts: dilemmas and challengesIntegrating social science expertise into AI teams raises real questions. How do you scope these roles? How do engineers and ethnographers communicate across different ways of producing knowledge? What happens when local insight conflicts with product timelines? We discuss what has been tried, what has failed, and where the friction points are. Building it into the companyThe expertise exists. The career pathways and organisational structures mostly do not. We discuss what it takes to create these roles and make them work in practice. Who Should Attend Speakers Speakers will be drawn from AI research, product development, and investment across China, East Africa, and South Asia. Interested in speaking? Contact us: [email protected] or [email protected] Organizers Olivia Yijian LiuCo-founder, Polyfoni. Business anthropologist and data scientist. Author of Start-up Wolf: The Shenzhen Model of High-Tech Entrepreneurship (Routledge, 2024). Mingyuan Zhang BetancourtUniversity of Oslo / Institute of Health and Society (HELSAM). Medical anthropologist researching health AI across China, Madagascar, and West Africa. About Polyfoni Polyfoni is an independent, Oslo-based nonprofit research communication platform. We center Global South social science in global AI discourse. Through editorial work, mentorship, events, and research partnerships, we connect the knowledge that social scientists produce in and about the Global South to the institutions and companies that need it. About Oslo Innovation Week This event aims to be part of the Oslo Innovation Week 2026 Side Programme. OIW brings together the Norwegian and international innovation ecosystem for a week of events, panels, and conversations. Check it out here.