Will AI replace Human Customer Service ?

The application of digital technologies, particularly AI, have continued to redefine the nature of retail and consumer services know Kai et al., (2020). One of the explored areas of interest is the use of AI- based chatbots that were integrated into online retail stores with an aim to provide timely interaction with customers (Cheng et al., 2021). The market for chatbots is set to experience a compound annual growth rate of 29.7% with the overall market reaching $125m by 2025 (Pantano and Pizzi, 2020). Nevertheless, despite the chatbot growth, they cannot fully satisfy the expectations of consumers based on their poor ability to identify the user’s necessities )Sheehan et al., 2020( and 70.2% of Facebook users have complained about failed interactions (Ashfaq et al., 2020). As such, people state that AI and human agents should work hand in hand to offer better service (Xiao and Kumar, 2021). Introducing chatbots into the mix but also maintaining humanity is the most critical aspect.

One limitation is that there is lack of effective information processing in human and computer interactions (Sheehan et al., 2020) and consumers don’t completely rely on the chatbot advice (Følstad et al., 2018) which influences the quality of the communication. Also, AI carries out data collection, storage and utilizing it in decision making for consumers and as consumers are aware that data gets leaked while usage of chatbots or any such AI models. Nevertheless, few studies proposed hypothesis on the relationship between service agent type, chatbot, or human, and communication quality and perceived privacy. Some research has been undertaken on consumer perceptions of chatbots, however, communication quality and privacy concerns regarding they have not been investigated adequately.

This study addresses this gap through a CASA design that analyzes the impact of a type of service agent on perceived communication quality and privacy risk in online retail. It also looks at how factors such as pressure to have face-to-face interaction with another person, moderates these factors and gives an insight on how AI and human agent can be used in e-commerce. This research suggests important implication for online sites and systems that wish to increase chatbot usage and augment human-computer interaction.