Firms that use AI to shape interactions with users report greater interconnected innovation, primarily because AI promotes knowledge integration rather than just information exchange. The innovation boost is substantially larger in firms with higher AI readiness.

Main Finding

AI-driven user–producer interaction (ADUPI) significantly increases user–producer interconnected innovation (UPII). This effect is transmitted through a dual knowledge mechanism—user–producer knowledge exchange (UPKE) and user–producer knowledge integration (UPKI)—with knowledge integration exerting a stronger mediating effect than knowledge exchange. AI readiness (AIR) strengthens the positive effects of ADUPI on both UPKE and UPKI, amplifying downstream innovation when firms are more AI-ready.

Key Points

• Core constructs
  • ADUPI: AI-enabled interactions between users and producers (e.g., personalized interfaces, recommendation/feedback loops).
  • UPII: Joint innovation outcomes emerging from sustained user–producer interaction.
  • UPKE: The transfer and sharing of knowledge between users and producers.
  • UPKI: The assimilation and recombination of exchanged knowledge into firm innovation processes.
  • AIR: Organizational preparedness (infrastructure, skills, processes) to exploit AI capabilities.
• Mechanisms
  • ADUPI → UPKE → UPII (knowledge exchange mediates part of the effect).
  • ADUPI → UPKI → UPII (knowledge integration mediates and has a stronger effect than exchange).
  • A joint (dual) mediation path: ADUPI influences UPII through both UPKE and UPKI.
• Moderation
  • AIR positively moderates ADUPI → UPKE and ADUPI → UPKI, meaning higher AI readiness amplifies both knowledge mechanisms and thereby innovation outcomes.
• Contributions
  • Shifts emphasis from firm-internal AI capabilities to cross-actor interactions as drivers of innovation.
  • Distinguishes between exchange and integration as separate and differently weighted knowledge processes.
  • Identifies AI readiness as a critical enabling condition for converting AI-enabled interactions into innovation.

Data & Methods

• Sample: Survey data from 974 firms.
• Analytical approach:
  • Regression analyses to estimate direct relationships.
  • Mediation analysis to test UPKE and UPKI as mediators of the ADUPI → UPII link.
  • Moderation analysis to test AIR as a moderator on ADUPI → UPKE/UPKI paths.
  • Bootstrap methods to assess the significance and robustness of mediation effects.
• Robustness: Results reported as robust across mediation/moderation specifications (details such as control variables and sectoral breakdowns not specified in the summary).

Implications for AI Economics

• Complementarities and heterogeneous returns
  • AI adoption benefits depend on complementary organizational capabilities (AIR) and interaction structures; returns to AI are heterogeneous across firms depending on their readiness and their ability to integrate user knowledge.
• Value of networked interaction
  • Economic value from AI is not only from internal automation but from enhanced user–producer interaction that creates knowledge spillovers and joint innovation—this shifts evaluation of AI investments toward ecosystem and interaction metrics.
• Policy implications
  • Policies that raise firms’ AI readiness (training, data infrastructure, standards) can increase social returns to AI by enabling firms to extract more innovation value from user interactions.
  • Support for knowledge-integration capacities (e.g., absorptive capacity, modular architectures, governance) may yield higher payoff than focusing solely on information exchange mechanisms.
• Managerial implications
  • Firms should invest in AI readiness (data pipelines, talent, governance) and explicit processes for integrating user-generated knowledge (product update cycles, cross-functional teams) to maximize innovation gains.
  • Design AI-mediated interfaces and feedback loops to facilitate both rich knowledge exchange and mechanisms to integrate that knowledge into development.
• Broader market effects
  • Firms with higher AIR may capture disproportionate innovation rents and accelerate product-market evolution, potentially influencing market concentration and diffusion dynamics.
• Directions for research/policy evaluation
  • When measuring economic impacts of AI, include interaction- and integration-based channels; consider longitudinal and causal designs to quantify dynamic effects and thresholds of AIR.