AI is reshaping sovereignty: control over data, computation and platforms is displacing territorial authority, creating infrastructural chokepoints that reproduce colonial patterns of extraction. Narrow regulations targeting applications will leave upstream power intact unless alternative, stewardship-based models of techno-sovereignty are supported.

Digital colonialism, techno-sovereignty, and infrastructural power

Artur Ishkhanyan · May 19, 2026 · Frontiers in Political Science

openalex theoretical n/a evidence 7/10 relevance DOI Source PDF

The paper argues that contemporary AI constitutes a form of digital colonialism that reconfigures sovereignty from territory to infrastructural and algorithmic control, and that governance focused only on downstream applications fails to address concentrated upstream infrastructural power.

This paper develops a political-theoretical critique of artificial intelligence (AI), situating contemporary AI infrastructures within debates on digital colonialism and techno-sovereignty. The analysis focuses on AI systems deployed for identification, classification, and governance—domains where sovereignty is most visibly reconfigured. It argues that AI constitutes a historically continuous yet technologically novel form of colonial power, shifting sovereignty from territorial authority toward infrastructural and algorithmic control—a transformation theorized here as infrastructural sovereignty. To substantiate this claim, the paper advances two arguments. First, it distinguishes digital colonialism from surveillance capitalism by showing how AI systems extend historical patterns of dispossession and epistemic domination beyond the commodification of individual behavior, embedding extractive and classificatory logics within data architectures, models, and standards. Second, it conceptualizes techno-sovereignty as a mode of authority grounded in control over data, computation, and AI infrastructures, exercised through state, corporate, and community or Indigenous configurations. Drawing on political theory, decolonial scholarship, and genealogical analysis, the paper traces continuities between colonial identification infrastructures and contemporary biometric and algorithmic systems, illustrating how sovereignty migrates from territory to systems, platforms, and protocols. Authority in AI systems, it shows, is exercised not through formal jurisdiction but through infrastructural chokepoints and dependency pathways that precede and condition law. It concludes by examining implications for AI governance, demonstrating that regulatory frameworks addressing only downstream applications leave the upstream concentration of infrastructural power largely intact, while community and Indigenous approaches—though constrained—offer alternative models of authority rooted in stewardship rather than extraction.

Summary

Main Finding

Contemporary AI infrastructures instantiate a technologically novel form of colonial power—digital colonialism—by shifting sovereignty from territorially grounded institutions to control over data, computation, and infrastructural chokepoints (what the author terms infrastructural sovereignty). This transformation is distinct from, though overlapping with, surveillance capitalism: it operates structurally and geopolitically (extraction, epistemic domination, dependency pathways) rather than primarily as commodification of individual behavior.

Key Points

Distinction: Surveillance capitalism focuses on behavioral commodification and individual harms; digital colonialism frames data extraction and AI as reproducing historical imperial patterns of dispossession, epistemic domination, and infrastructural dependency.
Genealogy: Modern AI identification/classification systems (biometrics, facial recognition, digital IDs, predictive scoring) are continuations of colonial identification infrastructures (fingerprinting, pass systems, registries) but transformed by automation, scale, and transnational integration.
Mechanisms of power: Three recurring mechanisms—(1) legibility via classification, (2) extraction via datafication of bodies/environments, (3) control via infrastructural mediation—explain how authority is reconfigured.
Infrastructural sovereignty & chokepoints: Sovereignty increasingly rests on control of chokepoints (semiconductor supply chains, cloud providers, compute capacity, identity rails, standards, proprietary models). Control is exercised materially (access, interoperability, pricing, updates) and often precedes or conditions law.
Typology of techno-sovereignty: Distinct but interacting forms—state techno-sovereignty, corporate techno-sovereignty, and community/Indigenous techno-sovereignty—produce different governance possibilities and constraints.
Dependency pathways: Upstream concentration (hardware, platforms, standards, model architectures) creates dependencies that reproduce digital colonial relations across applications and geographies.
Governance implication: Regulations focused only on downstream applications (privacy, use restrictions) often leave upstream infrastructural concentration intact. Community and Indigenous stewardship models offer alternative authority forms (stewardship vs extraction) but face material constraints.

Data & Methods

Approach: Interdisciplinary conceptual and genealogical analysis combining political theory, decolonial scholarship, and literature on data politics and AI governance.
Evidence base: Historical examples of colonial identification infrastructures; contemporary examples of biometric and AI systems; synthesis of existing scholarship (e.g., Couldry & Mejias on data colonialism, Zuboff on surveillance capitalism, Bratton on digital architectures, and empirical critiques of biometrics).
Methodology: Qualitative, theory-driven; develops conceptual distinctions, a typology of techno-sovereignty, and a sustained case study on biometric identification and infrastructural power.
Limits: Not an empirical econometric or large-scale quantitative study—findings are analytic and interpretive rather than statistical; relies on existing case studies and secondary literature.

Implications for AI Economics

Market structure and rents: Control of infrastructural chokepoints concentrates economic rents (compute, chip fabrication, cloud services, identity platforms). Economists should treat these chokepoints as sources of quasi-sovereign market power that shape returns to AI investment and firm strategy.
Barriers to entry & dependency costs: Upstream concentration raises fixed-cost barriers (specialized hardware, proprietary stacks, standards compliance) and creates dependency externalities for states and firms lacking infrastructural control—affecting industrial policy, trade, and comparative advantage in AI.
Redistribution and global inequality: Digital colonial dynamics imply value extraction flows that mimic historical center–periphery relations; economic analyses should incorporate cross-border distributional impacts, including non-market harms (epistemic loss, cultural exclusion).
Resource & externality accounting: AI infrastructure entails material and energy inputs (minerals, power). Economists should include environmental and supply-chain externalities in cost-benefit and policy models of AI deployment.
Policy levers and interventions: To alter economic incentives and reduce dependency, policies should target upstream structures: antitrust/enforcement on platform and semiconductor markets; strategic industrial policy for local compute and chip capacity; open standards and interoperable identity rails; support for publicly funded or community-controlled infrastructure; procurement rules that favor stewardship models.
Modeling suggestions for researchers: incorporate infrastructural chokepoints into IO and political-economy models; quantify dependency pathways (e.g., share of compute, vendor concentration, data pipeline ownership); evaluate welfare impacts of upstream vs downstream regulation; compare economic outcomes under state, corporate, and community techno-sovereignty regimes.
Evaluation of alternative models: Community and Indigenous stewardship can change value capture and governance incentives but require investments and protections (capacity building, legal recognition, funding). Economic analysis should estimate costs and long-term gains from decentralized or stewarded infrastructures versus concentrated commercial models.

Suggested immediate analytical priorities for AI economists: - Measure concentration at key chokepoints (chips, cloud, model providers) and map dependencies across countries/sectors. - Assess welfare and distributional consequences of infrastructural sovereignty (including non-market cultural/epistemic harms). - Model the effectiveness and trade-offs of upstream interventions (industrial policy, open-source/public infrastructure, antitrust) versus downstream regulation. - Incorporate lifecycle resource and energy costs of AI infrastructures into macro and sectoral policy analysis.

Assessment

Paper Typetheoretical Evidence Strengthn/a — The paper is a conceptual and political-theoretical analysis that does not present original empirical or causal identification; claims are supported by literature synthesis and genealogical argument rather than empirical testing. Methods Rigormedium — The paper uses careful conceptual distinctions, genealogical analysis, and engagement with decolonial and political theory literatures, which is appropriate and rigorous for a theoretical contribution; however, it lacks empirical validation, counterfactual testing, or systematic case-comparison that would raise methods rigor to high. SampleNo original dataset — a conceptual analysis drawing on political theory, decolonial scholarship, historical examples of identification infrastructures, and illustrative contemporary cases (e.g., biometric and algorithmic systems) rather than systematic empirical sampling. Themesgovernance inequality GeneralizabilityNo empirical generalization — claims are theoretical and require empirical testing across contexts, Focus on identification, classification, and governance domains may not extend to all AI applications (e.g., purely engineering or productivity tools), Genealogical and decolonial framing may reflect particular intellectual traditions and could be Western- or regionally-biased in its examples, Policy implications may vary across legal/jurisdictional regimes and technological ecosystems

Claims (8)

Claim	Direction	Confidence	Outcome	Details
AI constitutes a historically continuous yet technologically novel form of colonial power, shifting sovereignty from territorial authority toward infrastructural and algorithmic control (termed "infrastructural sovereignty"). Governance And Regulation	negative	high	configuration of sovereignty (territorial vs infrastructural/algorithmic)	0.12
AI systems deployed for identification, classification, and governance are the domains where sovereignty is most visibly reconfigured. Governance And Regulation	negative	high	degree of sovereignty reconfiguration in identification/classification/governance systems	0.12
Contemporary biometric and algorithmic systems show continuities with colonial identification infrastructures. Governance And Regulation	negative	high	continuity between colonial identification infrastructures and contemporary biometric/algorithmic systems	0.12
Digital colonialism is distinct from surveillance capitalism: AI extends historical patterns of dispossession and epistemic domination beyond the commodification of individual behavior by embedding extractive and classificatory logics within data architectures, models, and standards. Governance And Regulation	negative	high	degree to which AI architectures embed extractive/classificatory logics and reproduce dispossession/epistemic domination	0.12
Techno-sovereignty is a mode of authority grounded in control over data, computation, and AI infrastructures, exercised through state, corporate, and community or Indigenous configurations. Governance And Regulation	mixed	high	form and locus of authority over AI infrastructure (state, corporate, community/Indigenous)	0.12
Authority in AI systems is exercised not through formal jurisdiction but through infrastructural chokepoints and dependency pathways that precede and condition law. Governance And Regulation	negative	high	mechanisms of authority in AI systems (infrastructural chokepoints vs formal legal jurisdiction)	0.12
Regulatory frameworks that address only downstream applications leave the upstream concentration of infrastructural power largely intact. Governance And Regulation	negative	high	effectiveness of downstream-focused regulation in reducing upstream infrastructural power concentration	0.12
Community and Indigenous approaches offer alternative models of authority over AI infrastructure rooted in stewardship rather than extraction, although these approaches are constrained. Governance And Regulation	positive	high	viability and character of stewardship-based authority models for AI governance	0.06