A Dual-Channel Model of Artificial Intelligence Adoption Intention Among Executives of Traditional-Industry SMEs: Perceived Value, Perceived Implementation Risk, and the Boundary Condition of Digital Foundation

**A Dual-Channel Model of Artificial Intelligence Adoption Intention among Executives of Traditional-Industry SMEs: Perceived Value, Perceived Implementation Risk, and the Boundary Condition of Digital Foundation** **Weixiang Gan** Post Doctorate Researcher, Lincoln University College, 47301 Petaling Jaya, Malaysia **Qiaoran Liu** Agriculture and Rural Affairs Bureau of Yindu District, Anyang City 455000, China **Mengfei Xiao** Post Doctorate Researcher, Lincoln University College, 47301 Petaling Jaya, Malaysia…

By the Truvace deskOctober 18, 20262 min read1 sourcePeer-reviewedTRV-2026-1765· cite & verify

In brief

A survey of 526 executives of small and medium-sized firms in four traditional industries across four Chinese regions tested what drives their intention to adopt artificial intelligence. The authors fitted a structural equation model with two mediating pathways: perceived value and perceived implementation risk.

The result matters because the pathways pull in different directions. Competitive pressure and institutional support have total effects on intention that cannot be told apart from zero, so pressure and subsidies may not translate into adoption. The study is cross-sectional, so the causal order of the constructs remains unproven.

Main points

The sample was 526 SME executives in four traditional industries across four Chinese regions.
Performance expectancy and competitive pressure raise perceived value.
Effort expectancy and perceived institutional support leave value unchanged but lower perceived implementation risk.
Digital foundation weakens the suppressing effect of risk on intention, with the slope falling from -0.284 to -0.125.
The authors suggest building the data and systems base before the application layer.

The problem

Peer demonstration and subsidies do not move these executives, because the risk of implementing AI is one they cannot absorb, even where they see value in it.

The rundown

The study starts from the observation that such firms usually decide on AI through the chairman or general manager alone, with no information-systems department to consult. Earlier research mostly drew on firms that had already completed basic informatisation.

The model splits external conditions into two channels. Technology-side and environment-side factors act on perceived value or on perceived risk, and competitive pressure carries opposite signs across the two pathways, which cancels its total effect on intention.

Digital foundation acts as a boundary condition: it weakens the dampening effect of perceived risk on intention. The two product terms add only 0.007 to the explained variance in intention, a small increment.