Sovereign AI has quickly become one of the most repeated phrases in technology policy, and one of the least clearly defined. Does sovereignty mean national ownership, trusted access, regulatory control, sector-specific capability, data security, or resilience in a more fragmented world?  As AI becomes foundational infrastructure, the question is no longer simply who builds the best models. It is who controls the compute, data, energy, talent, capital and security architecture that modern economies will depend on.

Britain has committed to a twenty-fold increase in national compute capacity by 2030, launched AI Growth Zones, and begun investing in strategic capability through public infrastructure and research. But sovereign AI is not just an infrastructure challenge. It is also a question of geopolitical positioning, national security, industrial strategy and market design.

This session asks what sovereign AI really means in practice: what capabilities must be built at home, where international dependency is unavoidable, which sectors matter most, and how government, investors, regulators and industry should act now.  The sharper question is not whether sovereign AI matters. It is what kind of sovereignty is actually possible, what kind is desirable, and what choices in the next twelve months will shape Britain’s position by 2030.