Davos 2026 Debrief: AI competitiveness hinges on scale and sovereignty

As heads of state at the World Economic Forum in Davos last Tuesday focused on geopolitics, business leaders at the nearby AI House were speaking about separate yet interwoven imperatives: AI scalability and sovereignty.

Scalability refers to a common theme heard throughout the week at Davos 2026—it’s time to move beyond experimentation to the harder challenge of diffusion: How to scale AI safely, credibly, and profitably across organisations and economies. Sovereignty is no less challenging—protecting a country’s data and bolstering national competitiveness, at the risk of stifling cooperation and innovation across industries and markets.

To help frame these discussions, KPMG Germany launched the Strategic AI Capability Index (SACI) on Tuesday evening. The research, produced in collaboration with Oxford Economics and the German AI Association, quantifies how the US, Europe, and China compare across the AI value chain, from economic contribution and investment in skills and innovation, to regulation and public strategy shaping adoption. The SACI provides one of the first comprehensive assessments of how effectively major economies are positioned to develop, scale, and govern AI systems, and its findings underline a key point that resonated strongly in Davos: durable AI leadership depends on the integration of research strength, widespread economic adoption, and coherent strategic governance that reinforce one another over time. 

Our ongoing work on the economic impact of sovereign AI policies, presented to the ASEAN Business Advisory Council on the sidelines of the ASEAN Digital Ministers’ Meeting in Hanoi, supports our findings. It examines how different national approaches to AI sovereignty—and varying degrees of openness—shape adoption and productivity outcomes. Our analysis highlights the importance of strengthening domestic capabilities and resilience, while remaining sufficiently open to benefit from global innovation and cross-border technological diffusion.

Labour and skills availability underpin AI scalability. Previous research by Oxford Economics, undertaken with Cognizant, examined both the productivity upside and the adjustment costs associated with generative AI. Our New Work, New World study estimates that over the next decade most jobs in the US could experience some degree of disruption, with around 9% of the current workforce potentially displaced. At the same time, generative AI could add between US $477 billion and US $1 trillion to the US economy by 2032, depending on the pace of adoption. Conversations in Davos closely echoed these findings. Productivity gains are achievable, but they are not automatic. They hinge on how firms adopt AI, how workers transition, and whether trust can be built. 

Finally, the gathering reinforced a recurring message from our research: enterprise execution matters as much as frontier innovation. The next phase of AI will be defined by scaling—by translating cutting-edge innovation into systems that amplify human agency, align with societal values, and deliver broad-based economic gains.