This study aims to investigate the effects of using AI chatbots in Korean English education from a macro perspective. For this purpose, 19 experimental studies are selected to conduct a meta-analysis, synthesizing the results of 51 individual study cases. The results of this study are as follows: First, it is found that the overall effect size of using chatbots is more than medium size meaning that a chatbot is an effective tool to learn English. Second, in the aspects of linguistic competence and affective categories, each shows over medium sizes like the overall effect size. In details of the dependent variables, vocabulary and speaking in linguistic competence and motivation in affective categories, large effect sizes are shown. Third, the effect sizes are getting larger, as the younger the students are, the longer the experiment period lasts, and the more purpose-built the chatbot is. But the differences in the effect sizes in terms of these moderators (e.g., school level, experiment period, and chatbot type) are not significant. Lastly, it is suggested that follow-up studies are needed to collect a sufficient number of experimental study cases and subdivide the variables for performing a more detailed meta-anlysis.

This study compares AI PengTalk’s assessments of Korean children’s pronunciation with the assessments of Korean teachers. Sixty Korean sixth-graders participated as assessees, and four Korean elementary teachers participated as assessors. Both PengTalk and the teachers rated the children’s production of 10 English sentences on a five-point scale. They focused on segmentals, stress-rhythm, intonation, and speech rate. The findings were as follows: Firstly, PengTalk evaluated the children’s pronunciation in the four elements significantly lower than the teachers across all English proficiency levels. Secondly, teachers’ ratings of the students aligned more closely with their pre-evaluated English proficiency levels than the AI PengTalk’s assessments. The teachers rated students at the upper level significantly higher than those at the intermediate level, who were, in turn, assessed significantly higher than those at the lower level in all four elements. Furthermore, AI PengTalk and the teachers differed in the mean order of the four elements, particularly in segmentals. Based on the results of this study, suggestions were made for the development and implementation of AI-based English programs.

AI recommendation service is adopted in consumption consulting such as high-tech and fashion consumption (Thapliyal & Ahuj, 2021). Now, for high-tech and fashion products, the advance selling strategy is widely adopted. Thus, this study targets to detective the consumers’ preference toward AI agents comparing human agents under advance selling and spot selling. The independent variable of this study is consumption type: Pre-sale Products vs. Spot Products. Pre-sale Products are quite popular currently, especially technological products. Construal-level theory (CLT) offers a valuable framework to explain the mechanisms that trigger evaluations, predictions, and behaviors by linking the degree of mental abstraction (the construal level) to psychological distance (Trope & Liberman, 2000; 2003; 2010). Four dimensions including temporal, special, social, and probability distance are argued to present the psychological distance (Trope et al., 2007). Liberman et al. (2022) discuss the time distance and argue the distant-future events are represented in a more abstract, structured, high-level manner than near-future events. Kim & Duhachek (2020) draw on a dimension of persuasion by AI agents to posit that AI agents are perceived as low-construal agents because of the fact that people hold a lay theory that AI agents do not have superordinate goals and cannot learn from their experiences or possess consciousness like humans do. Consequently, they find that individuals perceive greater appropriateness and are more persuaded when an AI agent’s persuasive messages highlight low-construal as opposed to high-construal features. Moreover, consumers prefer abstract information related to a certain product rather than concrete information when a purchase is to take place in the distant future or when construal levels are high (Hernandez et al., 2015). Thus, this research hypothesizes: When consumers buy pre-sale products (vs. spot products), human agents will be the more favorable service provider than AI agents since the consumer is under a high level of construal. This research proposes to adopt a 2 (Advance Selling vs. Spot Selling) x 2 (Short Psychological Distance vs. Far Psychological Distance) x 2 (AI Agents vs. Human Agents) between groups experimental study to test the main effects and mechanism (H1). Furthermore, this study would identify the key moderating effects to discuss the boundary effects of the mechanism for establishing marketing strategies with AI services for managers.

With the dramatic advance in artificial intelligence, the use of an AI anchor in the news industry is a subject of great interest. An AI anchor is a computer-generated news anchor that mimics the human voice, appearance, and facial expressions to present the news. As the focus of this study is to investigate how the news media utilize an AI anchor in the news program, the research questions are as follows: (1) What are the differences in news topics between AI and human anchors? (2) How does the viewership differ based on these topic choices, with a focus on the emotional impact?

Despite the orientation towards online retailing journey accelerated by the application of new-age technologies in the pandemic context, the role of the physical store still has a central role in luxury shopping in the digital omni-channel perspective. Digital technologies have increased their impact on consumers (Evanschitzky et al., 2020; Klaus & Zaichkowsky, 2020; Kaplan & Haenlein, 2020; Davenport et al, 2020; Huang and Rust, 2021a; Pantano et al, 2022). In today’s digital age, AI is one of the new-age technologies raising growing interest for their potential disruptive impact on marketing and retailing in different sectors (Forbes, 2022).

As interactive marketing devices that serve as proximity-marketing tools, AI-powered voice assistants (VA) provide consumers with highly innovative convenience, which in turn fosters consumer–brand relationships (Wang, 2021). This research aims to explore the role of AI-powered VAs as a positive technology that offers consumers a sense of positive experiences, thus contributing to building a consumer-brand relationship. Based on the positive technology paradigm and transformation of flow strategy, this research conducted a 2 (locus of agency: high, machine-centric vs. low, human-centric) by 2 (brand image and voice congruency effect: incongruent vs. congruent) between-subjects experimental design. Then, ANOVA and structural equation modeling (SEM) analysis were conducted to explain how perceived control, flow, and happiness induced by the interaction with brands’ AI-powered VAs lead to the formation of brand loyalty under the moderating influences of brand image and VA’s voice congruency. A total of 316 participants were recruited via Prolific. The ANOVA analysis highlights the importance of user-centric agency, as people tend to desire to control their environment (White, 1959). Further, the results suggest that the congruency between brand image and VAs also leads people to positive reactions, as it improves their comfort and control (Rodero, Larrea, & Vázquez, 2013). SEM analysis results found that perceived control was a crucial factor that led participants to flow experience (Ghani et al., 1991). Further, this study found that perceived control could lead to a much broader aspect, such as an increase in happiness. Therefore, the overall study findings support the potential of AI-powered VAs as a positive technology. This research contributes to human-machine interaction, positive technology paradigm, and VA literature. In addition, this study provides beneficial insights for marketers and app developers.

Generative artificial intelligence (AI) tools such as ChatGPT, Dall-E, and Steve AI can facilitate instant content creation. As an illustration, a fashion brand can simply command ChatGPT to: “Write an Instagram caption about the importance of good clothes during winter in two hundred words”. In seconds, ChatGPT generates the output. This minimum effort and short production time in the usage of generative AI may enhance content marketing outcomes. However—as warned by scholars (Bruyn et al., 2020)—not only is it advantageous for businesses, AI can also be disastrous. Underpinned by this, and the fact that generative AI and content marketing are nonexistent in the pertinent literature (see streams of research on AI in marketing e.g., Eriksson et al., 2020; Huang & Rust, 2021, 2022; Vlačić et al., 2022), this research aims to explore the potential benefits and drawbacks of generative AI for content marketing.

Emotion AI, a subset of AI that measures, understands, responds to, and elicits human emotions, is an emerging area that has great potential for advertising research and practice. Studies on the applicability of emotion AI in advertising and marketing have been growing in academic journals. This rapidly burgeoning scholarship creates a need for advertising scholars to comprehend the current status of the research on emotion AI in advertising as well as opportunities and challenges that this new technological development will bring to. Thus, this study aims to offer an overview of research on emotion AI in advertising to identify the scope of existing research, gaps in knowledge, and opportunities and challenges that lie ahead.

Service encounters increasingly feature AI-powered inputs such as add-ons recommendations or aftercare solutions. These novel forms of customer service, provided by AI rather than humans, can shape customers’ sense of agency throughout the customer journey. Customers find themselves in a form of competitive collaboration with AI, sharing tasks, resources, inputs, and decisions. This research conceptualises and develops a scale to measure shared agency power during customer-AI interactions. Understanding the role of agency in AI- customer interactions is important, as agency represents a source, mechanism, delimiter and effect of a human’s or a machine’s actions. Agency may differ across various service encounters and with it, the type of perceived risks associated with human-AI interactions. Future research may use the shared agency power scale to better understand the nature and impact of customer-AI interactions in a service context on traditional marketing factors.

Artificial intelligence generated content (AIGC) refers to content produced by artificial intelligence that represents the perspectives of its users, and a new technique of content Generation. Continuous development in deep learning and algorithms have facilitated the adoption of AIGC. This research summarizes literature published under the topic of AIGC using bibliometric analysis method, aims to provide insightful research directions for future studies. 342 documents were collected from Database of Web of science, network visualization analysis among authors and citation analysis over publications are presented to scholars who wish to further research into this area.

Artificial intelligence (AI) is producing more and more branded content such as image, text, video and sound. This area of so-called generative AI became particularly popular with the public after the launch of ChatGPT. Furthermore, political correctness has been discussed in recent years, since society is becoming increasingly sensitive to certain issues surrounding topics such as racism or gender equality. Therefore, it is more important than ever for brands to communicate in a politically correct way. In the past, humans were responsible for negative brand communication and brand voice. However, with the development of these AI-tools and platforms, AI also creates brand voice and this AI-generated brand voice can similarly cause such negative feelings.

In addition to humanoid and robotic designs, an increasing number of AI-powered services are being represented by non-human species (i.e., zoonotic design). Yet, little is known about the consequential effects of such zoonotic AI on consumer adoption of these services. Drawing on the concept of speciesism and Cognitive Load Theory, the current research seeks to uncover how does using zoonotic (vs. robotic) designs affects consumer adoption.

AI technology has been increasingly integrated into a wider range of industries – from small-sized home appliances robots to service robots in the hospitality or education sectors. Subsequently, researchers are increasingly interested in understanding how consumers respond to AI robots. Adding to such stream of studies, this research delves into how consumer responses differ depending on the characteristics of AI robots, specifically, the degree of appearance resemblance to a real-life object and the benefit type that the robot is designed to provide. Particularly, this research focuses on the role of perceived fit in shaping consumers’ intention to adopt AI robots as the underlying mechanism.

Chatbot-based services in online travel agency (OTA) are rapidly spreading in order to respond more agilely to consumers' needs based on the digitalization of the travel industry. Although AI chatbots use anthropomorphism to provide social experiences on behalf of humans, research results on its effects are mixed. Therefore, based on construal level theory, this study suggests the degree of anthropomorphism (low vs. high) of chatbots prime mental representations of different construal levels (low vs. high) and the fit between anthropomorphism and communication context (communication types and conversation types) has a positive effect on use behavior. This research method consisted of sentimental analysis for exploring use behavior of AI chatbots and two experimental studies (study 1 and study 2) to examine the hypotheses. The results of this study expand construal level theory and avatar research to provide an understanding of the anthropomorphism of AI chatbots.

The following is not a conversation with a bank clerk. " Instead, let me introduce you to customized credit loans," "Do you want me to connect you to the screen of using COVID-19 support funds and checking balance?" These are the contents of consultations with AI chatbots at financial institutions. Chatbot, which used to be an additional tool for adding convenience to life, is now at the center of our lives.

Following a series of major breakthroughs in artificial intelligence (AI) technology, it is believed that the use of AI technology can fundamentally subvert many industries and business fields, one of which is marketing. For instance, AI is likely to become a key driver of how advertising and marketing activities are conducted (Qin and Jiang, 2019) and thus dramatically change marketing strategies and customer behaviors (Davenport et al., 2020).

The present study is designed to assess the ethical-moral effectiveness of three levels of AI implementations in the hospitality industry, which include mechanical AI for transactional services (automated service), thinking AI for functional services (algorithms search), and feeling AI for hedonic services (biometric sensors) when compared to consumers’ interactions with human. AI service robots are ethically challenging in use and morally controversial in its acceptance of labor replacement in hospitality contexts. In the hospitality industry, service robots have been rapidly adopted replacing frontline human services (Park et al., 2021). Service robots refer to “system-based autonomous and adaptable interfaces that interact, communicate and deliver service to an organization’s customers” (Wirtz et al., 2018, p.909). On the one hand, the applications of artificial intelligence (AI)-based service robots are promising in this field owing to remarkable accuracy in error reduction, portion control, and cost control in service operation and delivery (Berezina et al., 2019). However, on the other hand, there has been a debate on ethical and moral principles and values regarding service robots replacing human labor (Cowls et al., 2021). Nevertheless, restaurants’ adoption of service robots seems inevitable in the current marketplace as labor shortages and rising wages have challenged them to invest in automation (Tanzi, 2021). While prior research focused on the benefits of AI-based service offerings (e.g., Cristou et al., 2023; Huang & Rust, 2021; Park et al., 2021), this study explored the extent to which AI-based service robots are accepted by consumers without rising concern about service robots replacing human labors. To this end, we adopted Huang and Rust’s framework that identifies three levels of AI applications: mechanical, thinking, and feeling AIs. Mechanical AIs refer to the automation of repetitive and routine tasks (e.g., self-service technologies); thinking AIs facilitate rational decision-making based on data processing (e.g., conversational intelligent systems such as Siri); feeling AIs are able to interact with human emotions (e.g., humanlike robots that respond to human emotions such as Sophia). Further, we adopted the construal level theory (Trope & Liberman, 2003) to examine how different levels of AIs’ service capabilities influence the way that people think about AI-based service robots. In brief, this study demonstrated how different levels of AI benefits influence consumers’ moral concerns about AI-based service robotization’s replacement of human labor and social acceptance.

In this article, we address this shortcoming by exploring the concept of AI-based sustainable service—an offering that embeds artificial intelligence in ways that meet the needs of current consumers by contributing to socio-economic equalities and conserving the natural environment.

This paper aims to examine the effect of AI framing (i.e., Scientific AI vs Magic AI) on consumers' product evaluation. This study shows that magically framed AI technology may be more beneficial to appeal to product innovativeness when subjective product properties (e.g., personal taste) becomes important. On the contrary, when objective product properties (e.g., functionality) becomes important, scientifically framed AI technology is more likely to generate higher perceived product attractiveness and purchase intention.