The Sovereign AI Paradox: Why Owning the Infrastructure Is Not the Same as Controlling the Technology - a workload-level framework for control in a globally interdependent stack
Description
This working paper argues that sovereignty should be assessed for a workload and its full service path under a stated disruption, not inferred from asset location or ownership. It proposes a five-part diagnostic (control, continuity, substitutability, data authority, economic viability) with two cross-cutting requirements, an ordinal five-rung evidence ladder from announced to independently stress-tested capacity, and a non-compensatory decision rule. A targeted, non-systematic review of 2024–2026 sources illustrates the framework with six country cases (United States, China, France, Canada, India, United Arab Emirates). The framework is a proposed diagnostic, not a validated index; three hypotheses and a staged validation agenda are set out. Not peer reviewed. Appendix B records the verification status of each reference.
Files
Sovereign_AI_Paradox_Zenodo.pdf
Additional details
Related works
- Is derived from
- Other: https://medium.com/@alaa_9620/the-sovereign-ai-paradox-owning-the-infrastructure-is-not-the-same-as-controlling-the-technology-cea690934728 (URL)
- References
- Preprint: arXiv:2410.17481 (arXiv)
- Preprint: arXiv:2606.00729 (arXiv)
- Journal article: 10.1007/s00146-026-03116-4 (DOI)
- Journal article: 10.1007/s10676-024-09810-5 (DOI)
Dates
- Available
-
2026-10-27