Digital dependency: Big tech's control of platforms, data flows and infrastructure is recreating colonial-style power asymmetries that lock peripheral economies into extractive relationships; policy-led data sovereignty and alternative platforms provide important but uneven means of resistance.

Main Finding

Across two decades (2005–2025) of scholarship, dependency theory remains a powerful, updated lens for the digital age: a new form of "digital dependency" has emerged in which platform capitalism, data colonialism, and algorithmic control perpetuate and deepen historical core–periphery power imbalances. Digital inclusion without structural change can reproduce—and often intensify—extractive economic relations between multinational tech actors and countries or communities in the Global South.

Key Points

• Core concepts and mechanisms

  • Digital dependency: dependence on foreign-owned platforms, services, and standards that extract value from data and attention flows.
  • Platform capitalism: market power concentrated in a few multinational tech firms that capture rents through network effects, data aggregation, and two-sided markets.
  • Data colonialism: framing data extraction and monetization as a new colonial modality—resources are harvested from peripheral contexts and valorized elsewhere (Couldry & Mejias, 2019; Kwet, 2019).
  • Algorithmic control: predictive systems and automated decision-making reinforce unequal access, labor precarity, and surveillance regimes.
  • Technological lock-in: standards, APIs, and infrastructure choices produce path dependency that raises switching costs for users, firms, and states.

• Economic dimensions

  • Intangible data flows become a central source of value capture, complicating traditional trade and capital-account measures.
  • Financial neo-colonialism via fintech and digital financial services can channel rents and capital flows away from local economies.
  • Concentration of compute, model development, and cloud services centralizes AI capabilities in the core.

• Empirical manifestations

  • Dependence on foreign-owned digital infrastructure (cloud, undersea cables, app ecosystems).
  • Local ecosystems shaped by platform terms, advertising markets, and third-party SDKs that extract data and revenue.
  • Uneven development of AI capabilities: scarcity of local high-performance compute, talent drain, and uneven R&D investment.

• Counter-movements and agency

  • Digital/data sovereignty initiatives: national strategies, data localization, and legal reforms to assert control over data and infrastructure.
  • Indigenous/communal data governance models (Taylor & Kukutai, 2016).
  • Development of alternative platforms, open-source AI, and regional cloud/telecom initiatives (Hummel et al., 2021).
  • Policy innovation in the Global South seeks to rebalance bargaining power and create local value capture.

• Critical syntheses and debates

  • Neo-dependency frameworks adapt classical dependency ideas to intangible flows and platform-mediated value chains.
  • Ongoing debate about trade-offs: localization versus fragmentation, sovereignty versus innovation, regulation versus market access.
  • Calls to decolonize tech scholarship and move beyond techno-deterministic narratives.

Data & Methods

• Review type and scope

  • Systematic literature review and theoretical synthesis covering scholarship published 2005–2025 across political economy, STS, development studies, and critical data studies.
  • Sources include conceptual/theoretical works, qualitative case studies, comparative policy analyses, and an emerging set of quantitative studies (e.g., platform market share, data flow proxies).

• Common empirical approaches in the reviewed literature

  • Qualitative case studies (country/region-level policy responses, indigenous governance cases).
  • Policy and legal analysis (data protection, localization, antitrust).
  • Mapping of digital infrastructure ownership (cloud providers, undersea cables, data centers).
  • Bibliometric and discourse analyses in critical scholarship.
  • Limited econometric work measuring cross-border data flows, rent capture, or causal impacts of platform entry—an identified methodological gap.

• Limitations noted in the literature

  • Sparse comparable quantitative measures for data value flows and digital rent extraction.
  • Regional coverage uneven (some regions and sectors under-studied).
  • Difficulty attributing macroeconomic outcomes directly to digital dependency due to complex, interacting factors.
  • Rapid technological change outpaces many longitudinal studies.

Implications for AI Economics

• Market structure and rents

  • AI intensifies winner-takes-most dynamics: concentration of compute, models, and datasets magnifies platform-led rent extraction.
  • Data-as-capital implies cross-border value capture that traditional GDP/trade metrics understate; tax and redistributive systems may fail to capture AI-generated rents.

• Labor, productivity, and inequality

  • Algorithmic platforms mediate labor markets (gig work, crowdwork), often shifting bargaining power and compressing wages in peripheral contexts.
  • AI deployment can increase productivity but may worsen skill- and capital-driven inequality between core developers/owners and peripheral users/providers.

• Investment, diffusion, and lock-in

  • High fixed costs of AI infrastructure (data centers, specialized hardware) favor centralized providers and create lock-in for adopting economies.
  • Barriers to local AI development (compute, data access, talent) constrain technology diffusion and local value creation.

• Finance and monetary flows

  • Fintech and algorithmic credit can create new channels for capital extraction and dependence, including surveillance-backed lending and cross-border capital capture.
  • Stablecoins, payment platforms, and data-enabled credit models may shift financial sovereignty.

• Policy and governance

  • Antitrust and competition policy must adapt to data- and AI-specific market failures (platform gatekeeping, data monopolies).
  • Data governance regimes (localization, data trusts, interoperable standards) can reshape value capture if coupled with industrial strategy and investment in capabilities.
  • International coordination needed to address cross-border externalities, taxation of digital rents, and fair data flows.

• Research & measurement needs for AI economics

  • Develop quantitative metrics for data value flows, AI rent extraction, and compute concentration.
  • Causal studies on how platform/AI entry affects local firms, labor markets, and balance-of-payments.
  • Model the macroeconomic effects of data-driven capital flows, including scenarios for policy interventions (data taxes, investment in public compute).
  • Evaluate alternative ownership/governance arrangements (public cloud, cooperatives, open models) for improving distributional outcomes.

• Practical policy levers suggested by the literature

  • Invest in local AI infrastructure, workforce development, and R&D linkages.
  • Implement data-sharing frameworks and interoperable standards that reduce lock-in and enable local innovation.
  • Update taxation and competition frameworks to capture digital rents and curb anti-competitive practices.
  • Support indigenous and community-led data governance as part of inclusive AI strategies.

Concluding note The literature synthesizes a clear thesis: digital transformation reshapes dependency rather than erasing it. For AI economics, that means paying attention to data, compute, and platform power as central economic inputs whose ownership and governance determine where AI-generated value accrues. Future work should build rigorous measurement, causal inference, and policy evaluation into analyses to inform interventions that can decolonize digital development.