Firms that signal stronger data-driven decision-making in their disclosures tend to perform better internationally; the gains operate through sustainability-focused value-creation channels and are larger in competitive markets and in firms with significant foreign or state ownership.

Main Finding

Data-driven decision-making (DDDM), measured using AI language models (BERT and ChatGLM2-6B), has a positive effect on international firm performance. This effect operates through sustainable-value-creation mechanisms and is stronger in more competitive markets and in firms with higher foreign shareholding or state ownership.

Key Points

• DDDM is empirically linked to higher international firm performance.
• AI language models (BERT and ChatGLM2-6B) are used to quantify firms’ DDDM practices from textual sources.
• A theoretical/empirical framework identifies four value-creation pathways through which DDDM boosts international performance:
  • Current internal: pollution prevention
  • Future internal: green innovation
  • Current external: sustainability information disclosure
  • Future external: sustainability vision co-creation
• The positive DDDM → international performance relationship is amplified by:
  • Higher market competition
  • Greater foreign shareholding
  • State ownership

Data & Methods

• DDDM measurement: textual/semantic analysis of firm disclosures or managerial texts operationalized using transformer-based language models (BERT and ChatGLM2-6B) to construct a DDDM score.
• Empirical strategy: statistical/econometric analysis linking the DDDM measure to firm-level international performance, with mediation tests for the four mechanisms and interaction tests for the moderators (market competition, foreign shareholding, state ownership).
• Identification: the study combines AI-driven measurement with standard regression and mediation/moderation approaches to uncover mechanisms and heterogeneity (details not provided in the summary).

Implications for AI Economics

• Measurement innovation: Transformer language models can produce scalable, granular measures of managerial practices (like DDDM), enabling new empirical work on intangible capabilities.
• Value of AI-enabled capabilities: Firms’ use of data and AI-driven decision processes appears to create sustainable value that translates into better international performance—highlighting a channel by which AI adoption affects firm competitiveness and internationalization.
• Distributional effects and market structure: Amplification by competition and ownership structures implies that benefits of DDDM may be unevenly distributed across firms and institutional contexts; policies and investment decisions should account for these heterogeneities.
• Policy and managerial relevance: Encouraging data infrastructure, training, and transparency in sustainability-related practices may magnify the international payoff of DDDM, especially in competitive industries and mixed-ownership environments.
• Research directions: Use of LMs for firm-level constructs invites further validation (robustness, cross-country generalizability, causal identification), exploration of costs/risks (mismeasurement, strategic disclosure), and assessment of complementarities between AI tools and organizational governance.