Q&A: Unravelling the AI sovereignty paradox

The countries investing the most in AI are not necessarily the ones best positioned to benefit from it. This leads to the question as to what truly determines national AI competitiveness?

Answering this is Prakhar Neema, VP of Research and Strategy at Draup. Neema also considers why trying to build every part of the AI ecosystem domestically could actually put countries at a disadvantage. Neema draws on a recent report from Draup and the Ethical AI Governance Group (EAIGG), “The AI Sovereignty Paradox.”

Digital Journal: When people talk about the global AI race right now, the focus is usually on who has the best models and the most computing power. What are we missing?

Prakhar Neema: The bigger question is what happens after you have the technology. Can you actually put it to work across businesses, government, and the broader economy? That takes more than a strong model. You need the right talent, infrastructure, and data, along with organizations that can adapt quickly enough to use them. Our analysis with EAIGG shows the countries that can bring those pieces together are going to be in a very different position from those that simply have access to AI.

DJ: What does the “AI sovereignty paradox” mean, and why might countries that try to build every part of the AI ecosystem themselves actually put themselves at a disadvantage?

Neema: The paradox is that trying to control every piece of the AI stack can actually slow a country down. Very few countries have the capital, talent, infrastructure, and technology to build everything themselves. Chasing complete independence can pull resources away from the areas where they already have an advantage. The report points to a different approach: know what you need to control, know where you can rely on partners, and invest heavily in the capabilities that make your economy distinct. Sovereignty doesn’t have to mean doing everything alone.

DJ: You looked at five very different countries in the report. What did you learn from comparing them?

Neema: That there isn’t one formula for AI leadership. In the U.S., the sheer amount of capital and computing power available creates an advantage, especially when it comes to pushing new AI tech forward. Israel looks very different. Its startup ecosystem has been particularly good at moving ideas out of the research world and into actual businesses. India has something else working in its favor: scale. It has a deep pool of technical talent, along with digital infrastructure that continues to expand. France brings strengths in research and public investment. Japan stands out for its industrial base and ability to apply AI within advanced industries. The larger takeaway is that countries don’t need to lead in every category. They need to understand where they have an edge and build around it.

DJ: Having access to AI is one thing. Having the people who can actually put it to work is another. As adoption grows, what kinds of talent will countries and companies need most?

Neema: The talent needs won’t be the same everywhere. Some countries already have large engineering and technical workforces. The harder part is having people who can take AI into an organization and figure out how it should actually be used. That could mean changing a process, changing a role, or deciding that AI isn’t the right answer in the first place. The countries that develop more of that kind of talent will have an easier time turning AI investment into something useful.

DJ: The report identifies a growing “AI deployment gap” between having access to AI and being able to translate it into real economic value. What is preventing organizations from closing that gap today?

Neema: Access isn’t really the hard part anymore. The harder part is what happens inside the organization once the tech is there. A pilot can be launched fairly quickly, but changing the way people work takes longer. Companies have to make decisions about jobs, processes, and where AI should actually be used. That’s where a lot of organizations are getting stuck.

DJ: If access to AI keeps getting easier, what will actually separate the countries and companies that benefit most from it?

Neema: I think the difference will be how quickly they can change around the technology. Two companies can have access to the same models and get very different results from them. The same is true for countries. Having the tech matters, but so does having the talent, infrastructure, and institutions to put it to work. That’s why we think the next phase of the AI race will be less about access and much more about what you’re able to do with it.