Paper Type
Complete
Paper Number
PACIS2026-1326
Description
Small and medium-sized enterprises (SMEs) face structural barriers to artificial intelligence (AI) adoption, including limited budgets, data scarcity, and skill gaps, that render dominant resource-intensive AI paradigms inaccessible. This paper proposes the Service Science Framework for Frugal AI (SSFFAI), a Design Science Research artifact that integrates frugal AI engineering with Service-Dominant Logic to support cost-effective AI adoption in SMEs. SSFFAI defines four interdependent dimensions: value, management, science, and technology that are linked through a Support-Use-Enable logic that connects strategic objectives, business service systems, and lightweight AI techniques to the creation of multidimensional value. A four-stage co-creation method operationalizes the framework into actionable implementation guidance. Three illustrative instantiations demonstrate its applicability across manufacturing, retail, and professional services contexts. Unlike descriptive adoption frameworks, SSFFAI provides prescriptive design knowledge that reframes SMEs as resource-integrating service systems capable of achieving sustainable AI-enabled innovation.
Recommended Citation
Le Dinh, Thang; Vu, Manh-Chiên; and Dao, Viet T., "Service Science Framework for Frugal AI: A Design Science Approach for AI Adoption in SMEs" (2026). PACIS 2026 Proceedings. 4.
https://aisel.aisnet.org/pacis2026/ai_ml/ai_ml/4
Service Science Framework for Frugal AI: A Design Science Approach for AI Adoption in SMEs
Small and medium-sized enterprises (SMEs) face structural barriers to artificial intelligence (AI) adoption, including limited budgets, data scarcity, and skill gaps, that render dominant resource-intensive AI paradigms inaccessible. This paper proposes the Service Science Framework for Frugal AI (SSFFAI), a Design Science Research artifact that integrates frugal AI engineering with Service-Dominant Logic to support cost-effective AI adoption in SMEs. SSFFAI defines four interdependent dimensions: value, management, science, and technology that are linked through a Support-Use-Enable logic that connects strategic objectives, business service systems, and lightweight AI techniques to the creation of multidimensional value. A four-stage co-creation method operationalizes the framework into actionable implementation guidance. Three illustrative instantiations demonstrate its applicability across manufacturing, retail, and professional services contexts. Unlike descriptive adoption frameworks, SSFFAI provides prescriptive design knowledge that reframes SMEs as resource-integrating service systems capable of achieving sustainable AI-enabled innovation.
Comments
01-AIML