Vloga umetne inteligence in digitalne kulture pri spodbujanju digitalnih inovacij

This doctoral dissertation examines how organizations generate digital innovation and how digital technologies, particularly artificial intelligence (AI), are translated into innovation outcomes across digital processes, digital products, and digital business models.The dissertation is organized as a set of interlocking studies that collectively develop a coherent, mechanism-based explanation of digital innovation, grounded in digital innovation scholarship and dynamic capabilities theory.It responds to a central challenge in the literature: although digital technologies are widely expected to enable innovation, existing research often remains fragmented across type domains and levels of analysis, and it frequently underspecifies the organizational mechanisms through which digital technologies produce innovation.To address this challenge, the dissertation advances the argument that digital innovation outcomes are differentiated but interconnected, and that organizations generate them through capability chains in which culture, human-AI collaboration, and organizational assimilation processes play a pivotal role.The first study develops a comprehensive synthesis of digital innovation research in management and business.While several reviews of digital innovation exist, many adopt narrow perspectives, focusing on selected technologies, outcomes, or theoretical lenses, which has resulted in siloed understanding of the phenomenon.In response, this study maps the intellectual structure of the digital innovation field and proposes an overarching multilevel framework based on a Contextual Conditions -Mechanisms -Outcomes logic.The framework clarifies how contextual conditions, such as organizational culture and resources, enable organizational mechanisms, such as technology assimilation and capability development, which in turn produce distinct but related digital innovation outcomes.By organizing prior research in this way, the study provides a structured foundation for cumulative theorizing and sets the stage for the empirical studies that follow.The second study develops and empirically tests a model explaining how digital culture and AI assimilation jointly foster digital innovation through an internal capability chain.Digital culture is conceptualized as an organizational context characterized by openness to experimentation, collaboration, learning, and data-driven decision-making.The model positions digital culture as a foundational condition that enables AI to move beyond isolated experimentation and become embedded in organizational routines through AI assimilation.AI assimilation is theorized as a central organizational mechanism that reflects the routinized, organization-wide use of AI rather than mere adoption.The study further argues that AI assimilation primarily drives digital process innovation, as AI reshapes workflows, decision processes, and coordination routines.These process changes, in turn, enable digital product innovation by improving knowledge flows, accelerating development cycles, and enhancing cross-functional integration.Empirical results support this sequential logic and demonstrate that the relationship between AI and digital product innovation is largely indirect, operating through digital process innovation.In doing so, the study explains not only whether AI matters for innovation, but how and why its effects unfold through internal organizational mechanisms.The third study extends the mechanism-based logic to digital business model innovation.This study examines how employee AI-related competence becomes translated into business model renewal.It challenges the assumption that skilled employees and AI investments automatically lead to strategic innovation and argues that the critical missing element is again an organizational mechanism that converts dispersed competence into scalable and coordinated use.AI assimilation is theorized as this mechanism, enabling firms to integrate employee capabilities into shared routines and value creation logics.Drawing on dynamic capabilities theory, the study further proposes that the translation of AI-related competence into digital business model innovation is shaped by boundary conditions, particularly digital leadership and AI-related financial resources.Digital leadership is conceptualized as an orchestration capability that aligns incentives, coordinates cross-unit integration, and transforms local AI initiatives into organization-wide practices.Empirical findings support the mediating role of AI assimilation and highlight the importance of leadership in enabling business model innovation.Through this analysis, the study connects micro-level human capabilities to macro-level strategic outcomes via a meso-level organizational mechanism.Taken together, the studies advance a unified argument.Digital innovation types are distinct yet connected, and organizations generate them through capability chains in which digital culture, AI assimilation, and human-AI collaboration shape whether AI becomes routinized, scalable, and innovation-producing.The dissertation contributes to digital innovation research by integrating product, process, and business model innovation into a coherent explanatory framework, and to dynamic capabilities theory by clarifying how capability reconfiguration under digitization produces different innovation outcomes.By emphasizing mechanisms rather than direct effects, the dissertation offers a theoretically grounded and empirically supported explanation of how organizations convert digital technologies into sustained digital innovation.During the preparation of this dissertation, I used generative AI tool ChatGPT for limited language assistance, including grammar correction, stylistic refinement, and improving the clarity of the writing.The intellectual content of the dissertation, including all ideas, theoretical reasoning, interpretations, analyses, and conclusions, is entirely my own.

Summary

Main Finding

Organizations produce distinct but interconnected forms of digital innovation (digital processes, digital products, digital business models) through capability chains in which digital culture, AI assimilation, and human–AI collaboration act as pivotal organizational mechanisms. AI rarely produces strategic innovation directly; its impact is mainly indirect — AI assimilation routinizes and scales AI use, driving process innovation, which then enables product innovation, while translating employee AI skills into business-model renewal requires AI assimilation plus enabling boundary conditions (digital leadership and AI-related financial resources).

Key Points

Conceptual contribution
- Presents a unified, mechanism-focused framework (Contextual Conditions – Mechanisms – Outcomes) integrating process, product, and business-model innovation.
- Clarifies and operationalizes "AI assimilation" as routinized, organization-wide embedding of AI (distinct from mere adoption/experimentation).
- Situates analysis in dynamic capabilities and sociomaterial perspectives: capability reconfiguration and human–AI ensembles are central to innovation outcomes.
Empirical findings (summarized)
- Digital culture (openness to experimentation, collaboration, learning, data-driven decision-making) is a foundational antecedent that enables AI to be assimilated.
- AI assimilation primarily drives digital process innovation (workflow, decision, and coordination changes).
- Digital process innovation mediates the effect of AI assimilation on digital product innovation — the AI → product link is largely indirect.
- Employee AI-related competence alone does not scale to strategic (business-model) innovation; AI assimilation mediates this translation.
- Boundary conditions matter: digital leadership (orchestration, cross-unit coordination) and AI-related financial resources strengthen the pathway from AI competence → AI assimilation → digital business-model innovation.
Practical takeaways
- Investments in AI technologies must be paired with cultural, organizational, and leadership changes to generate sustained innovation.
- Measuring AI adoption is insufficient for assessing innovation potential — measures of assimilation, routines, and managerial orchestration are crucial.
- Policies or firm programs that combine skill development, organizational change, leadership capacity, and financing will be more effective than technology subsidies alone.

Data & Methods

Multi-method dissertation with three interlocking studies:
Comprehensive review and synthesis - Bibliometric analysis to map the intellectual structure of digital innovation research. - Topic modeling to identify foundational and emerging themes. - Interpretative and integrative qualitative synthesis to build the multi-level Contextual Conditions–Mechanisms–Outcomes framework.
Empirical study on digital culture, AI assimilation, and product/process innovation - Survey-based firm-level data collection (organizational measures of digital culture, AI assimilation, innovation outcomes, controls). - Measurement validation and structural modeling (measurement model and structural model estimation). - Tests for mediation (sequential double mediation) and endogeneity assessments. - Sociomaterial + dynamic capabilities theoretical framing.
Empirical study on employee AI competence and business-model innovation - Survey measures of employee AI-related competence, AI assimilation, digital leadership, AI-related financial resources, and business-model innovation. - Mediation analysis (AI assimilation as mediator) and moderated mediation (leadership and resources as moderators). - Measurement validation and robustness checks (common method bias addressed; measurement and endogeneity procedures reported).
Analytical approaches
- Combination of bibliometrics, topic modeling, qualitative integration for literature synthesis.
- Structural equation modeling / path analysis for hypothesis testing (mediation and moderated-mediation frameworks).
- Endogeneity and common-method bias checks to strengthen causal claims within cross-sectional survey limits.

Implications for AI Economics

Rethink how AI inputs are modeled in productivity and innovation studies
- Treat AI not simply as a capital input but as a process-dependent, organizationally mediated factor. Macro/micro empirical models should incorporate mediating variables (assimilation, routines, digital culture) and complementarities with human skills and leadership.
- The returns to AI investments are conditional: without assimilation and enabling culture/leadership, measured productivity and innovation gains will be muted — this helps explain heterogeneity in firm-level AI returns.
Measurement and identification guidance
- Recommend augmenting firm-level datasets (surveys, administrative data) with indicators of AI assimilation, digital culture, managerial practices, and AI-specific financial commitments to better capture effective AI use.
- For causal inference, prioritize longitudinal/panel designs that track assimilation processes, and exploit quasi-experimental variation in policies, leadership changes, or funding to identify causal effects of assimilation on innovation and productivity.
Labor and skills economics
- Policies that focus only on upskilling workers in AI tools are necessary but insufficient; scaling employee competencies into strategic outcomes requires organizational integration (assimilation) and leadership. Complementary investments in management practices and coordination capabilities should be part of labor-market interventions.
- Models of skill complementarities should explicitly include organizational mechanisms that convert individual skills into firm-level productive use.
Industrial policy and public intervention
- Subsidies for AI hardware/software should be paired with programs that support organizational change: leadership training, change management, standards/interoperability, data infrastructure, and funding for assimilation activities.
- Public evaluation of AI policy effectiveness should measure assimilation and downstream innovation (process/product/business model) rather than adoption counts alone.
Directions for future economic research
- Empirically link micro-level capability chains to macro outcomes: how widespread differences in assimilation affect aggregate productivity growth and structural transformation.
- Study diffusion dynamics: how digital culture and leadership conditions influence adoption-to-assimilation rates across sectors and firm sizes.
- Explore general-equilibrium effects of firm-level business-model renewal induced by AI (market structure, competition, labor demand composition).

Overall, the dissertation argues that the economic impacts of AI depend critically on organizational mechanisms and complementary conditions; for economists, incorporating those mechanisms into models, measurements, and policy design is essential to accurately assess and accelerate AI-driven innovation.

Assessment

Paper Typecorrelational Evidence Strengthmedium — Empirical claims are grounded in strong theory and multi-study designs that test mediated mechanisms and moderators, which provides plausible causal narratives; however, the reliance on observational organizational measures (likely cross-sectional or survey-based) without experimental or quasi-experimental variation limits confident causal inference and raises concerns about reverse causality and common-method bias. Methods Rigormedium — The dissertation adopts a coherent, multi-study approach combining a systematic literature synthesis with empirical tests of clearly specified mechanisms and boundary conditions, which is methodologically sound; yet, the summary does not report advanced causal strategies (panel/IV/difference-in-differences), nor does it specify sample construction, measurement validation, or controls for endogeneity, which temper assessments of rigor. SampleNot specified in the provided summary; empirical studies (Study 2 and Study 3) appear to use firm- or business-unit-level organizational survey measures of digital culture, AI assimilation, employee AI competence, digital leadership, AI-related resources, and self-reported innovation outcomes (digital process, product, and business-model innovation); details on sample size, industry, country coverage, sampling method, and temporal design are not reported. Themesinnovation org_design human_ai_collab skills_training adoption IdentificationTheory-driven mediation and moderation analyses using observational organizational data: the dissertation tests mechanism-based causal chains (digital culture -> AI assimilation -> process/product/business-model innovation) via statistical mediation and boundary-condition (moderation) tests, supported by multilevel theory and robustness checks; no randomized or quasi-experimental identification is reported in the summary. GeneralizabilityUnclear sample frame and geographic/industry coverage — may not generalize across sectors, countries, or firm sizes, Likely based on self-reported organizational survey data, raising common-method and reporting biases, If cross-sectional, limited ability to generalize causal timing or long-run effects, Findings concern innovation outcomes (process/product/business-model) rather than direct economic metrics (productivity, employment, wages), limiting macroeconomic generalizability, Context-specific factors (regulation, maturity of AI use, market structure) may affect applicability

Claims (9)

Claim	Direction	Outcome	Confidence & Evidence	Details
I develop a comprehensive synthesis of digital innovation research in management and business and propose an overarching multilevel framework based on a Contextual Conditions - Mechanisms - Outcomes logic. Other	positive	conceptual clarity / framework development	Reading fidelity high Study strength medium	not reported 0.3
Digital culture (openness to experimentation, collaboration, learning, and data-driven decision-making) enables AI to move beyond isolated experimentation and become embedded in organizational routines through AI assimilation. Adoption Rate	positive	AI assimilation (routinized, organization-wide use of AI)	Reading fidelity high Study strength medium	not reported 0.3
AI assimilation primarily drives digital process innovation, as AI reshapes workflows, decision processes, and coordination routines. Organizational Efficiency	positive	digital process innovation (changes in workflows, decision processes, coordination routines)	Reading fidelity high Study strength medium	not reported 0.3
The relationship between AI and digital product innovation is largely indirect, operating through digital process innovation (i.e., AI -> process innovation -> product innovation). Innovation Output	positive	digital product innovation	Reading fidelity high Study strength medium	not reported 0.3
Employee AI-related competence is translated into digital business model innovation only when an organizational mechanism — AI assimilation — converts dispersed competence into scalable and coordinated use (AI assimilation mediates the competence -> business model innovation relationship). Innovation Output	positive	digital business model innovation	Reading fidelity high Study strength medium	not reported 0.3
The translation of AI-related competence into digital business model innovation is shaped by boundary conditions, particularly digital leadership and AI-related financial resources (these variables moderate the competence -> AI assimilation -> business model innovation pathway). Innovation Output	positive	digital business model innovation (extent of business model renewal)	Reading fidelity high Study strength medium	not reported 0.3
Digital innovation types (digital processes, digital products, digital business models) are distinct yet interconnected, and organizations generate them through capability chains in which digital culture, AI assimilation, and human-AI collaboration are pivotal. Innovation Output	positive	digital innovation outcomes (process, product, business model innovation)	Reading fidelity high Study strength medium	not reported 0.3
By emphasizing mechanisms rather than direct effects, the dissertation offers a theoretically grounded and empirically supported explanation of how organizations convert digital technologies into sustained digital innovation. Other	positive	explanatory power / theoretical contribution	Reading fidelity high Study strength speculative	not reported 0.05
During preparation of the dissertation, I used generative AI tool ChatGPT for limited language assistance (grammar correction, stylistic refinement, and improving clarity); the intellectual content is entirely the author's own. Other	null_result	use of generative AI for language editing	Reading fidelity high Study strength high	not reported 0.5

AI matters for innovation only when it becomes routine: firms with a digital culture that routinizes AI use convert AI investments into process, product and business-model innovation, while diffuse competence or isolated experimentation yields limited strategic payoff.