Rather than viewing the Global South as mere markets, the piece argues these regions can design the next AI governance order, anchored in sovereignty and local needs.

Across regions, a pattern emerges: Sarawak in Malaysia deploys federated learning on local infrastructure, Turkey launches Kumru, a home-grown Turkish-language LLM, Qatar develops Fanar for Arabic, and Africa advances Masakhane and Lelapa AI with locally tailored models.

Recent export controls and sanctions are nudging nations to build independent AI ecosystems, reshaping the landscape in the Global South, the Gulf, Africa, and beyond.

Anthropic’s Fable 5 and Mythos serve as a case study of how US-led restrictions can trigger market volatility and feed the view of AI as a strategic weapon.

The author, drawing on international experience, underscores a push for an inclusive, sovereign-inclusive approach to AI governance.

Sovereign AI is a rising trend: middle powers and Global South actors deploy locally governed models on domestic hardware with data sovereignty, often featuring multilingual or localized capabilities.

US AI policy that relies on chokepoints has backfired, with countries advancing domestic and sovereign AI capabilities in response.

Shifts in governance could redefine control over AI infrastructure, moving toward a multi-polar landscape where sovereignty and localized models matter as much as scale.

Europe and the Global South are urged to act as architects of the next AI order, not passive customers, with Geneva poised to host a decentralized, globally inclusive governance framework.