Supply‑chain AI is a tool of geopolitical power as well as efficiency: an "algorithmic sovereignty trilemma" forces states to trade off optimisation, data sovereignty and interoperability, driving techno‑bloc formation and asymmetric vulnerabilities. Policymakers should favour audit‑transparent, modular interoperability to reduce rent‑seeking and systemic fragility while managing legitimate sovereignty concerns.

Main Finding

AI-driven supply chain risk-management systems should be understood not merely as operational optimization tools but as instruments of algorithmic governance that materially redistribute geopolitical power. The paper introduces an "algorithmic sovereignty trilemma" showing that states cannot simultaneously maximise technical optimisation, data sovereignty, and cross‑bloc interoperability; this forces trade‑offs that shape techno‑bloc formation, asymmetric vulnerabilities, and the geopolitics of AI value chains.

Key Points

• Reconceptualisation: Treats supply‑chain AI, data analytics, and ML systems as sites of algorithmic governance where technical design choices (bias detection, data routing, security protocols) function as direct levers of state power and risk allocation.
• Theoretical framing: Integrates structural realism with power‑transition theory to move beyond aggregate capability metrics and focus on institutionalisation and systemic effects of algorithmic supply‑chain governance.
• Multidimensional determinants of leverage: Proposes four core variables that determine geopolitical leverage via algorithmic systems:
  • Algorithmic bias exposure
  • Data security posture
  • Structural dependency concentration
  • Regulatory fragmentation risk
• Algorithmic sovereignty trilemma: A formal conceptual construct that captures the incompatibility of maximising all three objectives — technical optimisation, data sovereignty, and cross‑bloc interoperability — simultaneously. The trilemma maps the inevitable trade‑offs states face during technological decoupling.
• Illustrative applications: Applies the framework to semiconductor and critical‑minerals supply chains in the US–China rivalry and situates analysis relative to:
  • EU AI Act (regulatory standardisation/constraints)
  • US export‑control practices (export restrictions as chokepoint policy)
  • China’s data‑security architecture (domestic controls and routing)
• Normative and policy stance: Argues for policy recalibration toward resilient, audit‑transparent oversight architectures designed to reduce asymmetric vulnerabilities and limit weaponisation of data flows; warns that securitisation of AI value chains encourages exclusive infrastructure partnerships (techno‑blocs).

Data & Methods

• Methodological approach: Conceptual and mixed‑methods synthesis combining:
  • Theoretical integration (structural realism + power‑transition theory)
  • Development of a multidimensional evaluation framework (four determinants listed above)
  • Qualitative, illustrative case studies of semiconductor and critical‑minerals supply chains
  • Regulatory/policy analysis comparing the EU AI Act, US export controls, and Chinese data‑security measures
• Analytical focus: Emphasis on institutionalisation and architectural properties of systems (bias detection, routing, cryptographic and audit capabilities) rather than large‑N quantitative capability tallies.
• Empirical orientation: Illustrative rather than exhaustive empirical measurement; framework is intended as operationalisable for future empirical work (e.g., measuring dependency concentration, mapping cross‑jurisdictional information routing, or quantifying regulatory fragmentation impacts).
• Limitations acknowledged: Lays out a conceptual framework and case illustrations rather than presenting broad causal identification with large datasets; intended to guide subsequent empirical and modelling efforts.

Implications for AI Economics

• Redistribution of economic rents and market power: Algorithmic governance of data flows and oversight architectures becomes a source of rent extraction and strategic advantage — increasing incentives for states and firms to internalise critical infrastructure and forge exclusive partnerships.
• Trade and fragmentation costs: Regulatory fragmentation and sovereignty‑driven routing increase transaction and compliance costs, raise barriers to trade in digital inputs, and can lower global welfare relative to an interoperable equilibrium.
• Endogenous supply‑chain formation: Algorithmic dependencies (bias mechanisms, auditability, routing) should be modelled as endogenous determinants of where production and control locate, not as exogenous technicalities.
• Risk and investment allocation: Firms and states will reallocate investment toward auditability, cryptographic controls, provenance tooling, and localised redundancy — shifting the composition of capital and raising fixed costs for new entrants.
• Political economy of standards: Standard‑setting becomes a core battleground; divergent standards heighten first‑mover advantages and increase costs of interoperability, encouraging formation of techno‑blocs with attendant market segmentation.
• Empirical and modelling priorities for researchers:
  • Incorporate the four framework variables (bias exposure, data security, dependency concentration, regulatory fragmentation) into trade and international political‑economy models.
  • Measure algorithmic dependencies via network analyses of data and control flows, provenance tags, trade and investment linkages, and changes around regulatory shocks (e.g., AI Act implementation, export control announcements).
  • Estimate welfare and distributional impacts of different governance architectures (centralised audit‑transparent systems vs sovereignty‑centric closed systems).
• Policy prescriptions with economic effects:
  • Promote resilient, audit‑transparent oversight architectures (reduces asymmetric vulnerabilities; may raise upfront investment but lowers systemic risk).
  • Seek modular interoperability standards and verifiable audit protocols that lower cross‑bloc transaction costs while preserving legitimate sovereignty concerns.
  • Design export‑control and data‑sovereignty measures to minimise escalation and avoid endogenous hardening of techno‑blocs that reduces global efficiency.
  • Support diversification and redundancy investments (public incentives) to mitigate concentration rents and systemic fragility.

Overall, the paper reframes AI in geopolitics as an issue of algorithmic governance embedded in value chains. For AI economics, this implies shifting attention from capability tallies to the economic consequences of architectural choices, regulatory fragmentation, and the strategic design of algorithmic oversight.