Why do manufacturing firms struggle with artificial intelligence?

Abstract

Purpose This study examines the key barriers hindering the adoption of industrial artificial intelligence (AI) in European manufacturing firms. It aims to classify these barriers and analyse their influence on adoption decisions, offering a nuanced understanding that can support managerial strategies for AI implementation. Design/methodology/approach Drawing on data from the 2022 European Manufacturing Survey (EMS), the study analyzes responses from 472 firms across Spain, Slovenia, Slovakia and Croatia. Factor analysis was used to group perceived barriers into three categories: lack of resources, environment/ecosystem limitations and lack of usefulness/fit. Logistic regression models with interaction terms assess the direct and moderating effects of these barriers on AI adoption. Findings The study identifies lack of resources - particularly financial capital, infrastructure and skilled personnel - as an absolute barrier that directly inhibits AI adoption. Lack of usefulness/fit only hinders adoption when resources are also lacking, forming a conditional absolute barrier. Conversely, environment/ecosystem limitations and data quality concerns emerge as relative barriers, affecting firms already on the adoption path. Leadership support mitigates environmental challenges and plays a crucial moderating role in successful implementation. Originality/value The study introduces a novel conceptual distinction between absolute and relative barriers to industrial AI adoption, offering practical insights into how manufacturing managers can navigate AI adoption by identifying which barriers require foundational resolution versus those that arise later in the implementation process. This distinction adds a new dimension to the understanding of AI adoption barriers, highlighting the varying degrees of impact depending on an organization's stage in the AI adoption process