Ibero‑American SMEs are likely to capture value from AI only when strategy, leadership, sensing and institutional links align; targeted managerial training, sensing subsidies and intermediary networks raise adoption and returns. The paper provides a structured framework—and eight testable propositions—for economists and policymakers to model diffusion, complementarities and heterogeneous returns among resource‑constrained firms.

Main Finding

An integrative framework explains how Ibero‑American SMEs overcome resource constraints to adopt AI: four interrelated drivers — strategic synchronization, leadership commitment, technology sensing, and institutional bridging — interact with organizational antecedents and contextual moderators through implementation processes to generate AI‑driven performance improvements. The framework synthesizes RBV, dynamic capabilities, and institutional theory and yields eight testable propositions linking capability development to firm outcomes.

Key Points

• Research focus: AI adoption in resource‑constrained Ibero‑American SMEs and how strategic theories explain diffusion and value capture.
• Theoretical integration: combines Resource‑Based View (RBV), Dynamic Capabilities, and Institutional Theory to explain both internal capability development and external enabling conditions.
• Four enabling drivers:
  • Strategic synchronization — aligning AI initiatives with firm strategy and resource priorities.
  • Leadership commitment — top management support and vision for AI adoption.
  • Technology sensing — capability to detect, interpret, and trial relevant AI technologies.
  • Institutional bridging — leveraging networks, regulations, and intermediaries to access resources and legitimacy.
• Mechanism: organizational antecedents (e.g., existing resources, routines) interact with contextual moderators (market dynamics, institutional strength) through implementation processes (pilots, scaling, learning) to produce performance outcomes.
• Contribution: extends strategic management theories to an emerging‑market SME context and provides actionable guidance for practitioners; proposes eight empirically testable propositions.

Data & Methods

• Methodological approach: structured narrative literature review following Torraco (2016) and Juntunen & Lehenkari (2021).
• Corpus: 72 articles from multiple databases covering 2015–2024.
• Analysis: thematic synthesis and systematic coding to identify recurring patterns, drivers, moderators, and proposed causal links.
• Scope & limits:
  • Focused on literature; no new primary empirical data collected.
  • Geographical emphasis on Ibero‑America — findings are contextualized for emerging‑market institutional environments.
  • Time window captures recent AI literature but may omit post‑2024 developments.

Implications for AI Economics

• Theory and modeling
  • Complementarities: Treat AI as a capital/machine‑learning input whose returns depend on managerial and sensing capabilities; include interaction terms in production‑function models (AI × capabilities).
  • Adoption threshold models: Institutional bridging and leadership shift fixed costs and adoption thresholds, altering diffusion dynamics in heterogeneous firms.
  • Dynamic investment: Dynamic capabilities imply firms face path‑dependent investment decisions; incorporate adjustment costs and learning curves in growth models.
• Empirical research directions
  • Outcome variables: productivity (TFP), profitability, exports, employment composition, process innovation rates.
  • Explanatory variables: AI adoption intensity (project counts, software spend), indices of strategic alignment, leadership commitment surveys, sensing activities (R&D, partnerships), and measures of institutional support.
  • Identification strategies: firm‑level panel models, difference‑in‑differences for policy/intervention rollouts, instrumental variables for endogenous adoption (e.g., distance to AI hubs), matched field experiments (training or subsidy pilots).
  • Microdata sources: business registries, firm surveys (WB, IADB), administrative tax/employment microdata, procurement/contracting records, platform usage logs.
• Policy and market implications
  • Policy levers matter: reducing coordination costs (institutional bridging), subsidizing sensing and pilot projects, and leadership/managerial training can raise adoption and returns to AI among SMEs.
  • Heterogeneous returns: policymakers should expect varied returns due to firm capabilities and local institutions — targeting complementary capabilities can be more cost‑effective than blanket subsidies for software/hardware.
  • Spillovers and competition: strengthening SME AI capabilities may shift competitive dynamics, induce product market reallocation, and generate local spillovers; antitrust and labor policies should anticipate reallocation effects.
• Practical implications for SME strategy
  • Low‑cost tactics: prioritize strategic synchronization (start with high‑value, small pilots), invest in leadership buy‑in, build sensing via partnerships, and use intermediaries (incubators, chambers) to bridge institutional gaps.
• Testable propositions (research examples)
  • P1: The productivity payoff from AI adoption is increasing in firms’ dynamic‑capability scores.
  • P2: Institutional support (subsidies, hubs) lowers the adoption cost and increases the adoption probability among resource‑constrained SMEs.
  • P3: Leadership commitment moderates the effect of AI pilot projects on firm‑level scaling and long‑run performance.
  • (Framework includes five additional linked propositions amenable to panel and quasi‑experimental tests.)

Overall, the framework implies economists modeling AI adoption should explicitly account for managerial and institutional complements, heterogeneity across SMEs, and dynamic learning when estimating returns, predicting diffusion, or designing policy interventions.