Generative artificial intelligence as a governance mechanism in Lean Six Sigma–based new product development: an action research study in an SME

Purpose This study aims to examine how generative artificial intelligence (AI) can be governed and embedded within Lean Six Sigma–based continuous improvement and new product development processes in a resource-constrained small- and medium-sized enterprise (SME) context. Design/methodology/approach An action research methodology was applied in an industrial case study within a Portuguese frozen-food SME. The intervention followed a structured continuous improvement logic aligned with the define–measure–analyse–improve–control framework and was integrated within existing Kaizen routines, plan–do–check–act cycles and stage-gate governance processes. Generative AI (ChatGPT-4) was deployed as a governance-aligned cognitive support tool through a structured prompt library, human validation mechanisms and traceable coordination artefacts. Findings The intervention was associated with improvements in quality performance and operational coordination, including reductions in defect rates and rejected product volume, shorter decision cycles and improved alignment between shop-floor problem-solving and stage-gate decision-making. These improvements were observed without additional investment in technological infrastructure or specialised AI resources. The findings suggest that generative AI can support structured problem–solving, knowledge integration and coordination when embedded within disciplined continuous improvement routines and subject to human validation. Research limitations/implications Findings are based on a single SME case and reflect an intervention-based design. This study proposes exploratory propositions to guide future research on AI-enabled continuous improvement and governance mechanisms in SME contexts. Practical implications This study presents a low-cost, replicable approach for SMEs to integrate generative AI into existing Lean Six Sigma practices through structured governance, without requiring specialised AI capabilities. Social implications By supporting structured problem-solving, knowledge reuse and cross-functional learning, the proposed approach has the potential to improve workforce-level learning and access to technical knowledge within SMEs. Originality/value This paper contributes to the Quality 4.0 and Lean Six Sigma literature by examining how generative AI can be embedded as a governance-aligned cognitive support mechanism within continuous improvement systems, linking Kaizen routines and structured improvement cycles with stage-gate processes in an SME setting.