An analysis on the role of artificial intelligence in green supply chains

This paper investigates the retailer’s artificial intelligence (AI) adoption strategies in the green supply chain involving a manufacturer and a retailer. We demonstrate that the decision to introduce AI is influenced by the retailer’s estimation of the consumers’ green preference (CGP) without AI as well as the unit adoption cost of AI. Specifically, irrespective of whether the retailer underestimates or overestimates the CGP without AI, as the estimation bias increases, the retailer becomes more inclined to adopt AI; however, an increase in the unit adoption cost will discourage adoption. Furthermore, we find that if the retailer underestimates the CGP without AI, adopting AI may negatively impact the profits of both the manufacturer and the supply chain, as well as the greenness level, while simultaneously enhancing social welfare. Conversely, if the CGP is overestimated, adopting AI can improve the manufacturer’s profit and the supply chain’s profit but decrease the greenness level and potentially harm social welfare. We extend the model by considering the prediction accuracy of AI, demonstrating that as the prediction accuracy increases, the retailer who underestimates the CGP without AI becomes more inclined to adopt AI; however, this may not hold under certain conditions if the CGP is overestimated.