The Tripartite Battle for AI Sovereignty

The modern digital economy is fragmenting as nations and private actors race to secure the nodes of the AI ecosystem. Power is no longer defined by borders alone, but by who controls the semiconductor supply chains, cloud platforms, and biometric systems necessary for artificial intelligence to function. This concentration is extreme: Nvidia currently holds 92 percent of the market for data-center GPUs, while TSMC produces up to 70 percent of the world’s advanced chips. These bottlenecks are not natural, but the result of deliberate investment strategies and intellectual property regimes that favor established powers.

The Three Layers of Control

State techno-sovereignty is defined by a nation's capacity to independently develop AI infrastructure, often through domestic chip production or national language models. Japan’s Rapidus project and the U.S. focus on AI mercantilism exemplify this shift. However, these states remain paradoxically tethered to corporate giants. Meanwhile, firms like OpenAI and Google exercise corporate techno-sovereignty, setting the de facto rules of the digital landscape through proprietary models and cloud dominance. This creates a cycle where governments purchase access to their own technological capabilities through private providers.

Indigenous techno-sovereignty offers a third path, rooted in self-determination rather than extraction. Frameworks like CARE and OCAP challenge the prevailing "open data" model, which often leads to the indiscriminate scraping of cultural materials. By prioritizing collective stewardship over corporate or state-led accumulation, projects like Te Hiku Media’s Te Reo Māori speech recognition model demonstrate that algorithmic sovereignty can be achieved through community-led governance. Addressing the current imbalance requires a structural move toward digital federalism, where data trusts and Indigenous rights are legally integrated into global AI treaties to ensure democratic legitimacy.