# AI adoption in E-commerce enterprises: Insights into current practices and future directions from an interview study

## TL;DR

This study explores how e-commerce businesses in China are adopting AI, revealing varied practices and challenges across different company sizes and contexts.

## Contribution

The study introduces a new theoretical perspective on AI adoption in e-commerce through a qualitative interview approach.

## Key findings

- AI applications span e-commerce operations but small businesses struggle with technical depth and customization.
- Human-machine collaboration remains dominant, with AI handling pre-processing and humans doing fine-tuning.
- Organizational capabilities and employee AI literacy are critical, while external policies lack effective guidance.

## Abstract

With the accelerated integration of artificial intelligence (AI) technology in digital commerce, e-commerce businesses are showing a diversified trend in its application across processes such as marketing, operations, and customer services. This article, based on the TOE (Technology-Organization-Environment) model and the TAM (Technology Acceptance Model) framework, employs a multi-case qualitative interview approach to conduct semi-structured interviews with different types of e-commerce enterprises in Anhui Province, China. It explores their practical implementation paths, feedback on effects, and future plans regarding AI technology adoption. The research findings are as follows: First, AI applications have covered the entire chain of e-commerce operations, including content generation, advertising placement, data analysis, customer service management, and logistics scheduling; however, small and micro enterprises still face significant limitations in technical depth and customization capabilities. Second, while the effects of AI applications are emerging, most processes continue to rely heavily on human collaboration and supervision, resulting in a human-machine collaboration model of “AI pre-processing + human fine-tuning.” Third, the organizational capabilities of enterprises and the AI literacy of employees are key to adoption, while the external policy environment has yet to provide effective guidance. Fourth, enterprises exhibit a stratified integration pattern in AI deployment, ranging from ‘platform-bound’ to ‘tool-augmented’ to ‘self-developed’ models, reflecting significant dynamic variations across different organizational contexts. This study also provides a new theoretical perspective and empirical evidence for understanding the technology adoption logic and digital transformation practices in the AI-driven e-commerce industry.