Systematic AI Adoption Through Governance Mechanisms : A design science research approach to integrated roadmap development

Abstract

Organizations are adopting AI and generative AI rapidly, yet existing guidance is split into two tracks: capability and maturity models describing technical progression, and governance frameworks specifying controls and compliance requirements. This separation creates practical problems, as pilots can scale without adequate safeguards or compliance work can advance without underlying capability development. Although individual frameworks address governance or maturity separately, systematic integration with evidence-based progression criteria is still missing, creating a risk that organizations either over-govern before they have learned or under-govern while they are scaling.

The thesis addresses this gap by designing and demonstrating a governance-integrated roadmap for organizational AI adoption. Following Design Science Research Methodology, it answers three questions: (1) What governance mechanisms are required for systematic organizational AI adoption? (2) What gaps exist in current AI adoption frameworks regarding these mechanisms? (3) How can these mechanisms be integrated into a stage-based roadmap that accommodates organizational maturity levels while ensuring evidence-based progression? A semi-systematic literature review on AI governance, maturity assessment, and adoption frameworks derives six governance mechanisms—Environmental Governance, Organizational Governance, AI-System Governance, Auditing & Standards, Maturity & Readiness, and Human–AI Agency. Comparative analysis shows that existing frameworks address these independently but rarely integrate them systematically. Empirical work in a low-maturity manufacturing SME, using semi-structured interviews, participant observation, and retrospective activity mapping, grounds the design requirements and provides an initial, single-case evaluation of applicability.

The designed artifact is a four-stage AI Adoption Roadmap—Foundation & Assessment, Experimentation & Learning, Operationalize & Integrate, and Enterprise Scaling—that operationalizes the six governance mechanisms across Technology Readiness Levels with explicit activities, outputs, and evidence-based stage-gates. A four-month single-case demonstration of Stage 1 in a low-maturity manufacturing SME resulted in seven design guidelines: integrate governance mechanisms systematically with technical progression; allow methodological flexibility in readiness assessment for resource-constrained organizations; use prototyping to discover hidden readiness gaps; apply evidence-based stage-gates to prevent premature advancement; treat document quality as a foundational capability for GenAI deployment; conduct use-case-specific regulatory analysis scaled to risk and exposure; and adopt augmentation-first patterns in early-stage AI adoption.

The research contributes a governance-integrated roadmap design translating abstract principles into actionable organizational activities, and empirical insights advancing understanding of early-stage adoption in low-maturity organizational contexts. It advances academic discussion by offering initial empirical support for integrating governance and maturity through evidence-based progression and by showing how this can be operationalized under resource constraints. Practically, the work provides guidance for organizations starting AI adoption and for actors supporting them. The study is limited to a single-case, Stage 1–only, four-month demonstration in one low-maturity manufacturing SME, without comparative evaluation across organizations or later stages. Future research should evaluate and refine the roadmap and guidelines through longitudinal studies over complete adoption lifecycles and comparative studies in different sectors and regulatory contexts.