The Challenge of Sovereign AI

The term “Sovereign AI” refers to efforts by a nation and the businesses operating within its borders to develop, deploy and govern AI under their own laws, in order to preserve control over data, models and other aspects of the development and implementation of AI tools. Sovereign AI’s goal is to avoid dependence on foreign vendors for critical AI capabilities.

This excerpt from a recent Deloitte report on “The State of AI in the Enterprise” highlights the significant implications that Sovereign AI has for companies operating in multiple jurisdictions:

The rise of sovereign AI has immediate practical implications. Companies working across borders must navigate complex requirements that vary by country, forcing them to build customized solutions for different markets. More than 3 in 4 companies (77%) now factor an AI solution’s country of origin into their vendor selection decisions, and nearly 3 in 5 (58%) now build their AI stacks primarily with local vendors. This signals that geographic sovereignty is now as important as innovation.

As the former vice president of observability at a major telecommunications company noted, “I’ve been working with a lot of international companies lately that are adamant we use an in country infrastructure. We’re taking the approach with some of those customers to do a distilled small language model for them that meets the import/export control rules because we build it in their country. With state-run companies in particular, there is skepticism when you’re using something from outside the country.”

Deloitte notes that Sovereign AI issues vary significantly based on geography and industry group. It points out that only 11% of companies in the Americas rely on foreign sourced solutions for a majority of their AI stack, but that nearly 1/3rd of EMEA companies do. However, US companies are increasingly required to take into account state-level deployment considerations.