Is AI sovereignty possible? Seeking a balance between autonomy and interdependence.

Key Chapter Title List

1. Introduction
2. Part I: What is AI Sovereignty?
3. Drivers of AI Sovereignty
4. How Countries are Responding to AI Sovereignty
5. Part II: Strategies for AI Sovereignty
6. Why Absolute AI Sovereignty is Unattainable
7. Why Choose Managed Interdependence
8. Putting Managed Interdependence into Practice
9. Conclusion
10. Appendix A: Categorical Data and Country Rankings
11. Appendix B: Detailed Description of the AI Technology Stack

Document Introduction

As artificial intelligence increasingly occupies a central position in global public policy and discourse, AI sovereignty has become part of the vocabulary for many policymakers. This concept bundles together strategic, economic, and cultural autonomy claims concerning critical infrastructure, data, and governance rules. Its concerns stem from multiple objectives, reflecting both legitimate government interests and potentially counterproductive considerations. AI is built on a global foundation—transnational research collaboration, complex supply chains, information technology networks, and vast data reflecting human knowledge and activity—no country can completely isolate itself from it. This report aims to explore how to understand and manage these interdependencies to achieve the legitimate goals of AI sovereignty.

The potential impact of AI development and its rapid proliferation have intensified digital sovereignty concerns worldwide and added an extra layer of urgency. The dominance of the United States and China in AI development and deployment, coupled with the geopolitical competition between these two global powers, is prompting other nations to seek to narrow the gap and avoid being caught in the middle. Countries' ambitions around AI computing power, data, and models take various forms, aiming to enhance security, resilience, economic competitiveness, and cultural-linguistic inclusivity through AI sovereignty strategies. As India, a leader in AI sovereignty initiatives, is set to host the AI Impact Summit in February 2026, this issue will take the international stage.

There are valid reasons why countries seek autonomy over AI systems. Supporting multiple languages undoubtedly enhances the utility of AI, allowing broader access to the knowledge and benefits it brings. Developing or operating AI systems domestically can yield social benefits and is often seen as necessary for national security and competition both at home and abroad. However, these benefits are not guaranteed; their complexity and cost may render them unfeasible or inefficient, and their performance, resilience, and security may fall short of international alternatives. Consequently, sovereign AI systems could lead to stranded investments or underutilization. Conversely, sovereign AI systems could also become tools of digital authoritarianism or serve as strategies for certain globally influential nations to consolidate or expand existing dominance.

The core argument of this report is that achieving full-stack AI sovereignty is structurally infeasible for almost all countries. AI is a transnational technology stack with highly concentrated chokepoints across minerals, energy, computing hardware, networks, digital infrastructure, data assets, models, applications, and cross-cutting enablers like talent and governance. Pursuing absolute sovereignty would inevitably lead to market fragmentation, divergent standards, and duplicative or ineffective public investment.

Based on this, the report proposes a pragmatic alternative path: managed interdependence. This approach does not pursue self-sufficiency but relies on strategic alliances and partnerships to mitigate risks across all layers of the AI technology stack. Countries can practice managed interdependence by mapping dependencies by technology stack layer, prioritizing feasible interventions, diversifying suppliers and partners, and embedding interoperability and portability through technical standards, procurement, and governance mechanisms. Implemented effectively, managed interdependence can enhance national resilience and capacity for autonomous action while preserving the benefits of open markets and cross-border collaboration.

The report concludes that AI sovereignty presents complex trade-offs and poses a series of critical questions to global AI actors, including the United States and China seeking to diffuse their AI products, and the many other countries desiring their own AI systems: How can the economic benefits of domestic AI systems be captured while avoiding inefficient investment, inadequate performance, and diminished competitiveness? How should countries reconcile AI sovereignty with international cooperation in areas like security? How can governments ensure sovereign AI systems protect human rights rather than becoming tools of digital authoritarianism? How should countries manage these objectives in ways that promote, rather than hinder, global AI development and governance? The answers to these questions will shape the landscape and stability of the future global AI ecosystem.