Technology, AI & the Future of Innovation
When we talk about reskilling or upskilling, we often assume that a system exists. Structured programs, employer support, some form of safety net. But retraining programs, unemployment insurance, and employer sponsored learning were built for formal employment relationships. In economies where informality dominates, they reach only a small share of workers. The rest are outside that system entirely. When AI exposure rises, it does not automatically translate into opportunity. It can deepen exclusion, because the support structures people would need to adapt are simply not there. The second is youth disconnection. Across the same regions, more than one in four young people between 15 and 24 are not in employment, education or training. This is not only a social issue. It is a signal about how workforce capacity is built over time. These are the individuals who will enter, or fail to enter, labor markets already being reshaped by AI. At the same time, AI is already starting to remove many of the entry level roles that once served as the first step into a career. Those early roles are where people learn how to work, build judgment, and understand how organizations function. When that step disappears, the effects are not immediate, but they build over time. The gap shows up later in the middle tier. The layer of experience that organizations depend on does not fully develop, because the pipeline was never built in the first place. What the data makes clear is that there is no single response to this transition. Countries are not starting from the same point, and they are not facing the same mix of pressures. Each one faces a different combination of exposure, readiness, labor market structure, and institutional constraint. AI will keep moving. The real question is whether countries understand their own position well enough to respond in the right way. In some places, the priority will be governance. In others, it will be skills, education, labor market protections, or stronger institutions. The response has to fit the risk. That is now the leadership challenge, not simply to promote AI, but to govern it with a clear understanding of national realities. The countries that manage this well will be the ones that understand where they stand, set the right priorities, and turn technological change into a deliberate source of workforce and economic advantage. Sources: Maldonado Valencia, C.A. (2026). “Global Labor Market Vulnerability to AI: Readiness, Risk and Inequality.” SSRN Working Paper No. 6251833. Exposure scores: ILO Working Paper No. 140 (2024), weighted by ILOSTAT Labour Force Statistics. Readiness: Oxford Insights Government AI Readiness Index (2024). Supplementary: Cerutti et al., IMF WP 25/76 (2025).
Countries with high exposure and high readiness, such as the United States, Singapore, Japan, Switzerland and Luxembourg, face a governance challenge. The institutions exist, adoption is underway, and the priority now is to build regulatory frameworks, workforce strategies, and accountability mechanisms, including audits and public reporting to keep pace with it. Countries with high exposure and low readiness, such as Bosnia and Herzegovina, Belarus, Cuba, and Montenegro, face a stability risk. Disruption arrives before the institutional architecture exists to manage it. The priority there is to stabilize first by building institutional capacity before adoption moves further ahead. Countries with low exposure and high readiness, such as India, China, Indonesia, Peru, and Thailand, still have a strategic window, but it is likely shorter than it appears. How they use that time will define their labor market position for the next decade. Countries with low exposure and low readiness, such as Haiti, Madagascar, El Salvador, and Sri Lanka, face the slowest moving but most foundational challenge. Without civil registration systems, digital infrastructure, and educational capacity, any AI specific policy risks becoming disconnected from the economy it is supposed to serve. The risk is not immediate disruption. It is that by the time AI becomes more relevant to these economies, the gap will have widened beyond the point where targeted policy can still close it. But quadrant position alone does not determine outcomes. Two structural conditions shape whether any transition policy can reach the workers it is meant to support. The first is informality. In Sub Saharan Africa, more than 84 percent of workers have no formal contract, no social protection, and no access to the systems that support job transitions. In South Asia, the figure is 77 percent; in Latin America and the Caribbean, 62 percent. These are not small groups. This is most of the workforce.
DAVOS CLUB 21
Powered by FlippingBook