avos 2026 was not short on insights, it was short on resolve. AI and the geopolitical landscape were woven into virtually every discussion, within the official World Economic Forum as well as the side events.
Sovereign AI, infrastructure, and governance emerged as the necessary anchor for AI independence. The messaging ranged from cautious optimism to fever–pitch warnings on the main stage while at side events, practitioners shared hard–edged implementation challenges.
The prevailing narrative was that sovereign AI and infrastructure will enable nations to secure their data and shield them from the demands of more powerful nations. There is truth in this. Yet sovereignty can be achieved without isolation. Nations do not need to build every layer of the AI stack from the ground up—most can’t. What matters is training models on sovereign data, not to compete, but to reflect the culture, language, and experience of those the technology is meant to serve.
Canadian Prime Minister Mark Carney’s rousing address, calling for the middle powers to unite around shared values, is logical in response to the U.S.–China competition over AI and economic supremacy. With this competition, we have normalized a race with no agreed destination. Supremacy in service of what exactly?
Carney’s speech served as a benchmark for the week, galvanizing geopolitical sentiments. However, we must consider that this strategy may unintentionally serve up a third superpower aiming for supremacy. As values shift, one unanswered question is how to organize around humanity over hegemony.
There were conversations about human–centric policy, shared prosperity, and the stewardship of technology. But the striking absence was any significant framework for moving those conversations into action. Anthropic CEO Dario Amodei put a clear point on it from the main stage, warning that AI may be “uniquely well–suited to autocracy.” This was a direct challenge to a world in which technology scales faster than governance.
With so many brilliant minds convened on one snowy mountain in Switzerland, Davos is unmatched in focusing our world view and centering our attention on what matters. Now the question is whether it can move beyond convening to committing to transparent, global agreements for 2027, starting with data sovereignty standards, workforce transition frameworks, and a shared definition of what human–centric AI actually requires in practice.
a global affairs media network
Can Davos move beyond insights, serve up commitment?
Image by via Unsplash+
March 4, 2026
After a month to reflect, members of our delegation to this year’s World Economic Forum shared their thoughts on what struck and stayed with them most. For Amy Peck, insights from Davos 2026 on sovereign AI were abundant, but what’s missing is a roadmap for implementation.
D
avos 2026 was not short on insights, it was short on resolve. AI and the geopolitical landscape were woven into virtually every discussion, within the official World Economic Forum as well as the side events.
Sovereign AI, infrastructure, and governance emerged as the necessary anchor for AI independence. The messaging ranged from cautious optimism to fever–pitch warnings on the main stage while at side events, practitioners shared hard–edged implementation challenges.
The prevailing narrative was that sovereign AI and infrastructure will enable nations to secure their data and shield them from the demands of more powerful nations. There is truth in this. Yet sovereignty can be achieved without isolation. Nations do not need to build every layer of the AI stack from the ground up—most can’t. What matters is training models on sovereign data, not to compete, but to reflect the culture, language, and experience of those the technology is meant to serve.
Canadian Prime Minister Mark Carney’s rousing address, calling for the middle powers to unite around shared values, is logical in response to the U.S.–China competition over AI and economic supremacy. With this competition, we have normalized a race with no agreed destination. Supremacy in service of what exactly?
Carney’s speech served as a benchmark for the week, galvanizing geopolitical sentiments. However, we must consider that this strategy may unintentionally serve up a third superpower aiming for supremacy. As values shift, one unanswered question is how to organize around humanity over hegemony.
There were conversations about human–centric policy, shared prosperity, and the stewardship of technology. But the striking absence was any significant framework for moving those conversations into action. Anthropic CEO Dario Amodei put a clear point on it from the main stage, warning that AI may be “uniquely well–suited to autocracy.” This was a direct challenge to a world in which technology scales faster than governance.
With so many brilliant minds convened on one snowy mountain in Switzerland, Davos is unmatched in focusing our world view and centering our attention on what matters. Now the question is whether it can move beyond convening to committing to transparent, global agreements for 2027, starting with data sovereignty standards, workforce transition frameworks, and a shared definition of what human–centric AI actually requires in practice.
