| Home | E-Submission | Sitemap | Editorial Office |  
top_img
Journal of Korean Society for Quality Management > Volume 52(3); 2024 > Article
업무 환경에서 생성형 AI 사용 의도에 영향을 미치는 촉진 요인과 저해 요인 분석

Abstract

Purpose

This study aims to investigate the factors influencing the adoption of Generative AI in the workplace, focusing on both enablers and inhibitors. By employing the dual factor theory, this research examines how knowledge support, customization, entertainment, perceived risk, realistic threat, and identity threat impact the intention to adopt Generative AI technologies such as ChatGPT.

Methods

Data were collected from 192 participants via MTurk, all of whom had experience using Generative AI. The survey was conducted in June 2024, and the data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) to ensure the validity and reliability of the measurement model. Attention-check questions were used to ensure data quality, and participants provided demographic information at the end of the survey.

Results

The findings reveal that knowledge support and entertainment significantly enhance the intention to adopt Generative AI, whereas realistic threat poses a substantial barrier. Customization, perceived risk, and identity threat did not significantly affect adoption intentions.

Conclusion

This study contributes to the literature by addressing the gap in understanding the adoption mechanisms of Generative AI in professional settings. It highlights the importance of promoting AI's knowledge support and entertainment capabilities while addressing employees' concerns about job security. Organizations should emphasize these benefits and proactively mitigate perceived threats to foster a positive reception of Generative AI technologies. The findings offer practical implications for enhancing user acceptance and provide a foundation for future research in this area.

1. Introduction

In era of advancing AI technologies such as ChatGPT, there has been a transformation in how businesses engage with customers and streamline various operations. These AI systems are adopt at generating text that closely mimics language, simplifying tasks like handling customer inquiries creating content and offering support. Particularly noteworthy is ChatGPT’s proficiency in engaging in conversations sharing information and aiding with writing tasks (Dwivedi et al., 2023). This technological progress has broadened the horizons for office-based industries by enabling them to leverage AI for customer service, content creation and administrative functions.
Nevertheless, the widespread adoption of AI models raises concerns and hurdles. Ethical dilemmas surrounding the use of AI generated content include risks associated with misinformation, dissemination, or misuse, well as the necessity for oversight have all gained prominence. For example, incidents such as the protests by the American Writers Guild highlight the growing apprehension about the implications of AI in creative industries (Isidore, 2023).
Consequently, organizations need to navigate these challenges when integrating AI models into work environments to ensure effective utilization. Hence, this study aims to explore the motives driving office workers utilization of Generative AI in their job responsibilities.
There is a growing interest in AI that can create content, with studies focusing mainly on its applications in education and only a few examining its impact in work settings. For example, Generative AI offers advantages in education by providing personalized instruction, adaptive learning, and automated assessments, which can deliver effective education to learners (Adıgüzel et al., 2023; Halaweh, 2023; Lim et al., 2023). These are contents that are difficult to provide effectively in traditional education due to a lack of human resources. However, at the same time, the use of AI in education raises ethical concerns, including potential biases, privacy issues, and over-reliance on technology (Adıgüzel et al., 2023). It could also signify the end of traditional forms of writing education, such as essays (Lim et al., 2023). Due to these dual characteristics, there is significant debate over the use of Generative AI in the field of education.
However, its importance in work settings is also notable, yet related research is lacking. Previous studies have discussed the impact of introducing Generative AI in the workplace and the changes required to maintain sustainable employability for workers (Adiasto, 2024). Ayinde et al. (2023) discuss the potential for Generative AI's adoption in the workplace. However, these studies are primarily theoretical, and empirical research remains limited.
To address this gap in research, our study applies the dual factor theory to identify two factors that influence people’s willingness to adopt AI in the workplace, enablers and inhibitors. Through surveys and data analysis, we seek to uncover the underlying motivations for both embracing and resisting AI in environments. This initial exploration of integrating AI into work environments aims to establish a foundation for research efforts, in this field.

2. Theoretical Background and literature review

2.1. Generative Ai in the workplace

The popularity of Generative AI has grown with the emergence of ChatGPT(Hussain et al., 2024). By integrating deep learning and language models, ChatGPT has greatly improved the capabilities of traditional chatbots. Its advanced language processing skills allow it to be effectively utilized in various workplace settings. One prominent application is in customer service, where ChatGPT's ability to understand and respond to human language in real time is expected to significantly enhance efficiency. Additionally, it excels in data analysis and report generation, going beyond mere text summarization to analyze data and create detailed reports. In fields like programming, ChatGPT demonstrates impressive problem-solving abilities.
Research on Generative AI has predominantly focused on its application in educational contexts. Studies have explored its potential to aid in scientific writing (Salvagno et al., 2023) and its impact on educational environments, debating whether it would be beneficial or detrimental (Lim et al., 2023). However, most of these studies remain conceptual, with empirical research still in its early stages. Specifically, there is a lack of research examining the role and acceptance of Generative AI in workplace settings.
With the rise of AI driven automation, like ChatGPT, there's a growing fear of job instability among workers (Cao & Song, 2024). Additionally, businesses might be wary of implementing Generative AI tools due to worries about sensitive data leaks. To highlight the importance of further investigation, the upcoming section will focus on the dual factor theory.

2.2. Dual factor theory

Dual factor theory divides the determinants of technology usage intentions into two categories: enablers and inhibitors (Cenfetelli, 2004). The theory was first introduced by Herzberg et al. (1959), who discovered that job satisfaction and dissatisfaction arise from different factors. Contrary to the common belief that satisfaction and dissatisfaction are merely opposite ends of a single spectrum, this theory asserts that they are influenced by distinct variables. Traditional models of technology acceptance, such as the Technology Acceptance Model (Davis, 1989) and the Unified Theory of Acceptance and Use of Technology (Williams et al., 2015), attempt to explain the adoption of new systems based on their perceived usefulness (Park et al., 2022). However, they often overlook inhibitors. Considering the crucial role that barriers to innovation play in technology adoption, both enablers and inhibitors must be examined (Park and Park, 2024; Park et al., 2024). Despite the growing recognition of Generative AI, particularly ChatGPT, its professional application remains limited. This study, therefore, employs Dual Factor Theory to explore and understand the factors that influence the intention to adopt Generative AI in the workplace.

3. Research model and hypotheses development

3.1. Enablers

3.1.1. Knowledge support

One of the valuable resources provided by advanced AI models is knowledge support, which brings significant benefits to office worker industries. In this context, knowledge refers to the information and insights obtained through these AI systems(Wang et al., 2023). Generative AI and similar models have the ability to analyze vast amounts of text data, making them indispensable tools for supporting knowledge acquisition. They can extract relevant insights from diverse sources such as articles, research papers and internal documents, giving employees access to a wealth of information.
By utilizing these AI models for knowledge support, employees can experience multiple advantages. First, they can quickly find and access relevant information, saving time and effort compared to manual searches. With the help of Generative AI models, employees can efficiently locate the information they need for their work tasks.
Moreover, Generative AI models can provide real time suggestions, recommendations and summaries based on analyzed data. These models can generate concise summaries that highlight key points and offer suggestions for further exploration. By incorporating these AI generated insights into their workflow, employees can make better decisions and improve their work quality.
Also, Generative AI can be valuable for fostering creativity and innovation. Workers have the opportunity to collaborate with these systems to brainstorm ideas, create content and compile reports. Generative AI's language generation features can assist employees in expressing their ideas more clearly, ultimately enhancing the quality and productivity of their written work.
For example, a study by Jo and Park (2023) highlighted the importance of Generative AI's ability to provide knowledge support in influencing users willingness to use the system. The study revealed that users find great value in gaining new knowledge through Generative AI, which positively impacts their actual usage of the platform. Based on these insights, following hypothesis is proposed:
H1: Knowledge support has a positive impact on the adoption intention of Generative AI in the workplace.

3.1.2. Customization

Another benefit of using AI technology is the ability to provide personalized experiences. Customization involves tailoring content, products and services to meet the specific needs of users (Xiao & Benbasat, 2007). AI systems driven by machine learning algorithms can personalize recommendations, solutions and interactions based on individual preferences and requirements. By analyzing extensive datasets and utilizing advanced algorithms, AI can gather and analyze user preferences, behavior patterns and contextual information. Notably, sophisticated natural language processing algorithms like ChatGPT can process unstructured data, such as text, to deliver outputs customized to meet specific user requirements and enhance interactions (Rudolph et al., 2023).
In white-collar industries, such customization can offer numerous benefits. For example, in customer service settings, AI driven chatbots and virtual assistants can offer customized responses to customer queries that based on their specific needs. By comprehending user intent and context, these systems can provide relevant support customized for each individual, thereby improving customer satisfaction levels. Likewise in marketing and sales tasks, AI technologies leverage customer data to offer personalized suggestions and recommendations. Through analyzing previous purchase histories, browsing behaviors and demographic details, AI powered systems can predict individual preferences accurately and suggest tailored product recommendations that boost customer engagement rates leading to increased conversions. Considering these findings, this study proposes the following hypothesis:
H2: Customization has a positive impact on the adoption intention of Generative AI in the workplace.

3.1.3. Entertainment

Entertainment is another critical antecedent of technology adoption. Entertainment plays a crucial role in attracting individuals to interact with information sharing platforms like the internet and social media (Cheung et al., 2021). While customers may utilize new technologies for their practicality, they also find joy and amusement in using these tools (Bianchi & Andrews, 2018). In particular, conversation stands out as a fundamental form of entertainment for humans. It fosters emotional connections, facilitates knowledge exchange and offers pleasure. Engaging in conversations allows people to express ideas, share jokes and discuss various subjects, helping them alleviate everyday stress and expand their knowledge, thus, emphasizing the importance of conversational technologies like chatbots from an entertainment standpoint.
Several studies have highlighted the entertainment aspect of chatbots. For example, Ashfaq et al. (2020) discovered that the entertainment factor provided by chatbots significantly impacts users intention to continue using them. Similarly, Kasilingam (2020) proposed that entertainment plays a pivotal role in driving the use of chatbots in shopping scenarios. Building on these earlier findings, this research proposes the following hypothesis:
H3: Entertainment has a positive impact on the adoption intention of Generative AI in the workplace.

3.2. Inhibitors

3.2.1. Perceived risks

One of the job demands associated with the adoption of Generative AI models in workplace is privacy concern. As AI technologies handle and process large amounts of data, including personal and sensitive information, there are potential risks and challenges related to privacy protection (Shin, 2021). Privacy risks are defined as the negative impact resulting from the disclosure of users' personal information (Bouhia et al., 2022). Privacy encompasses both personal and corporate information. Traditional research has primarily focused on the infringement of personal data privacy. Customers fear that companies might misuse their personal identification or credit card information, leading to reluctance in using certain services. Privacy breaches have been a major barrier to the adoption of online services.
In the context of Generative AI, however, the potential for privacy invasion arises from the AI learning from the input data provided by users. Given Generative AI's problem-solving capabilities, users might input extensive work-related data to seek solutions, which can lead to privacy breaches (Song et al., 2022). Consequently, workers must continuously verify whether it is safe to use such data with Generative AI in their workplace. This continuous verification process is perceived as an additional task, which can act as a strain in using Generative AI. Considering these factors, we propose the following hypothesis:
H4: Privacy risks have a negative impact on the adoption intention of Generative AI in the workplace.

3.2.2. Perceived threat

The rise of new technologies brings about enhanced efficiency and advantages for people. However, some technologies also present threats to humans. A notable example is robots. When individual perceived certain threat, they tend to avoid the threat (Yoo and Park, 2023). Robots are perceived as a threat to human jobs, safety, and resources, particularly when they are capable of outperforming humans on various tasks (Yogeeswaran et al., 2016). The concern posed by robots escalates when they adopt more human like appearances rather than purely mechanical forms (Yogeeswaran et al., 2016). Although Generative AI lacks the physical embodiment of robots, it exhibits a significant linguistic resemblance to humans. This linguistic similarity could lead Generative AI to be perceived as human-like, potentially raising similar concerns as those associated with humanoid robots.
The perceived threats can be broadly classified into realistic threats and identity threats. Realistic threats encompass worries related to physical safety, job stability and material possessions (Huang et al, 2021).
For example, the more people worry about job security, the more they tend to resist using technology (Dabbous et al., 2022; Cao and Song, 2024). Therefore, the notion of Generative AI as a threat can instill apprehension in individuals who may fear losing their jobs or facing changes in their socioeconomic status.
Identity threats emerge when individuals feel that the uniqueness, values, or identity of their group is being questioned or undermined (Huang et al., 2021). In this study, identity threats are based on the definition by Craig et al. (2019), which suggests that when AI contradicts an individual's personal experiences or identity, the individual may experience a loss of self-esteem and engage in defensive behaviors to maintain their identity-related self-worth.
In the realm of Generative AI, particularly those designed to mimic human traits, there is a concern that these AI models could potentially erode the distinction between humans and machines, posing a risk to human individuality and distinctiveness. These perceived threats may trigger feelings of unease and unfavorable perceptions towards Generative AI. Research in the hospitality context supports this assumption, showing that higher levels of identity threats posed by AI systems lead to lower intentions to use them (Xu et al., 2024). Considering these factors, we propose the following hypothesis:
H5: The realistic threat of Generative AI has a negative impact on the adoption intention of Generative AI in the workplace.
H6: The identity threat of Generative AI has a negative impact on the adoption intention of Generative AI in the workplace.

4. Methodology

The conceptual framework for this research is grounded in dual factor theory. To comprehensively examine both the enablers and inhibitors influencing the intention to use Generative AI in the workplace, we have identified three enablers (knowledge support, customization, and entertainment) and three inhibitors (perceived risk, realistic threat, and identity threat) as independent variables. This study aims to understand the factors driving the adoption of Generative AI in professional settings, and therefore utilizes a survey methodology for data collection. The research model is depicted in Figure 1.

4.1. Survey Items

Respondents were asked to address six variables that represent the benefits and risks associated with Generative AI. The benefits include knowledge support, customization, and entertainment, while the risks encompass perceived risks, realistic threat, and identity threat. To ensure the reliability and validity of the measurements, items were adopted from previous studies.
Knowledge support was measured with three items from Wang et al. (2023). Customization was assessed using three items from Cheung et al. (2020), and entertainment was evaluated using three items from Li et al. (2013) and Wolf and Maier (2024). Perceived risk was measured with three items from Lu et al. (2011). Realistic threat and identity threat were each measured with two items from Huang et al. (2021). The intention to use Generative AI was measured with three items from Lin (2011). All items were rated on a seven-point Likert scale (1 = strongly disagree, 7 = strongly agree) to ensure high reliability. The complete list of measurement items is provided in Table 1.

4.2. Data Collection

This study surveyed users experienced with Generative AI, collecting data from June 3 to June 5, 2024. A total of 210 office worker participants from the United States were recruited via MTurk, with strict criteria to ensure data reliability (Aguinis et al., 2021). Participants were required to have a 98% or higher approval rating and have completed over 500 tasks on MTurk. Only those meeting these standards were permitted to participate, and each received a $0.50 reward.
To further ensure data integrity, attention-check questions (e.g., "Please select 'Strongly disagree' for this question") were included. Respondents who failed these checks were excluded from the study. At the survey's conclusion, demographic information was gathered from all participants. Table 2 presents an overview of these demographics. After excluding 18 responses that did not pass the attention tests or provided incorrect answers to reverse-coded items, 192 valid responses were retained for analysis.

5. Results

The data were analyzed using the partial least squares method (PLS) and the Smart PLS (version 4.0) statistical software package. This method is when analyzing small samples and measurement scales because of its low constraints (Chin 1998; Gefen et al. 2011). Additionally, PLS allows latent constructs as formative or reflective indicators (Fu 2011; Hsieh et al. 2012). This enables analysis of the second order variables used in this study. Considering the varying user experiences, AI Policy was included in the model as a control variable.

5.1. Measurement validity test

In this study, Smart PLS 4.0 was utilized to conduct data analysis through Partial Least Squares Structural Equation Modelling (PLS-SEM). The choice of PLS-SEM was strategic, given its robustness in handling non-normal data, a prevalent challenge in survey-based research (Hair et al., 2019). Additionally, PLS-SEM is particularly advantageous for modeling intricate structures with smaller sample sizes (Hair et al., 2019)
To ensure the reliability of our analysis, we rigorously tested the validity of the measurement items, focusing on both convergent and discriminant validity. Following the guidelines of Fornell and Larcker (1981), convergent validity was evaluated based on three criteria. Composite Reliability (CR) and Cronbach's alpha values should exceed 0.7, and an Average Variance Extracted (AVE) should be greater than 0.5.
The analysis revealed that all factor loading values ranged from 0.772 to 0.963, meeting the required criteria. The Cronbach's alpha values for all items were robust, with the lowest value being 0.777. Additionally, the lowest AVE recorded was 0.769, well above the minimum threshold of 0.5. Therefore, the convergent validity of the measurement model was confirmed, as detailed in Table 3.
For assessing discriminant validity, we employed the Heterotrait-Monotrait Ratio (HTMT) method, as recommended by Hair et al. (2019). Discriminant validity is established when HTMT values are below 0.9. As illustrated in Table 4, all variables satisfied this stringent criterion, thereby confirming the discriminant validity of the measurement model.

5.2. Hypotheses test

Next, we tested our hypotheses. Initially, we examined the R-square score, which stands at 0.429, indicating that our research model explains 42.9% of the variance in the intention to adopt Generative AI. We then applied the bootstrap method with 5000 samples for robust hypothesis testing. The results indicate that among the three enablers, knowledge support (β=0.411,p=0.019) and entertainment(β =0.400, p=0.026) have a significant impact on adopt intention. However, customization does not have a significant effect on adoption intention (p=0.513). Therefore, hypotheses H1 and H3 are supported while H2 is rejected.
Next, among the three inhibitors, only realistic threat has a significant impact on adoption intention (β=-0.187, p=0.019). Thus, H5 is supported. Contrary to our expectations, perceived risk (p=0.778), and identity threat (p=0.878), do not have a significant impact on adoption intention. Therefore, H4 and H6 are rejected. According to Cohen (1988), all adopted relationships showed a weak effect (f2 ≥ 0.02). Table 5 presents a summary of the results.

5. Discussion

This study provides important insights into how Generative AI can be effectively utilized in workplace settings. The findings indicate that knowledge support and entertainment are key factors in encouraging the adoption of Generative AI at work. Specifically, Generative AI's ability to assist with tasks such as automating repetitive processes, creating macros, and providing instant access to information directly enhances productivity and efficiency in the workplace. These advantages reduce the cognitive load on employees, allowing them to focus on more complex tasks, which can significantly improve overall job performance. This finding aligns with previous research emphasizing the importance of knowledge support functionalities in technology adoption (Jo & Park, 2023).
In contrast, customization was found to have no significant impact on the intention to use Generative AI. This finding contrasts with previous research in the context of chatbots, where customization has been shown to positively influence usage intention (Lee and Park, 2019). Prior studies have predominantly focused on the consumer perspective, where customization may be considered important. However, in workplace settings, factors such as efficiency, reliability, and accuracy are often prioritized over customization. This context may explain why knowledge support, which directly aids in work-related tasks, significantly influences usage intention, while customization does not.
Entertainment represents an extended feature of Generative AI. Traditional information systems operate on predefined rules, constraining users to specific questions and tasks. In contrast, Generative AI uses natural language processing to understand and respond to human communication without such constraints. The ability to engage in interactive and entertaining communication positively influences the inclination to use chatbots, aligning with previous studies (Ashfaq et al., 2020; Kasilingam, 2020).
The results also indicate that perceived risk did not significantly affect the adoption of Generative AI. Although there are concerns about potential risks, such as the leakage of personal information or corporate secrets, users seem to mitigate these risks by avoiding the direct input of highly sensitive information. Instead, they may opt to ask indirect questions or use AI for routine tasks, which helps in reducing security concerns.
Notably, the research revealed that the perception of realistic threats creates reservations about incorporating Generative AI into work tasks. Unlike previous technologies perceived as aids to human endeavors, AI advancements are regarded as potential job disruptors. This apprehension contributes to a reluctance to embrace generative AI in professional settings. These conclusions echo earlier studies suggesting that robots can be viewed as threats by colleagues.
The study by Xu et al. (2024) argues that both realistic threat and identity threat negatively impact the intention to use AI, with realistic threat being more significant. The findings of this study support these conclusions. Identity threat has been identified as a significant concern in research related to robots (Huang et al., 2021; Singh et al., 2021). These studies found that as robots appear more mechanical, the identity threat is lower due to the clear distinction from humans, whereas humanoid robots, which resemble humans, tend to evoke a higher level of identity threat. While Generative AI mimics human language, it lacks physical characteristics, unlike robots. Considering this, it can be inferred that identity threat is more influenced by visual cues rather than linguistic communication cues.

6. Conclusion

The rise of Generative AI has brought about significant changes in society, including the workplace. Despite these shifts, there are still uncertainties surrounding how these new technologies are embraced or rejected in workplace settings. This study aims to address this gap by investigating both the enablers behind and inhibitors to adopting this advanced technology, drawing on the dual factor theory. The findings of this research provide several implications.

6.1. Theoretical implication

First, it is significant as an initial exploration of how Generative AI is being used in workplaces. While the use of Generative AI in real world scenarios is on the rise, its application at work is mostly limited to a few areas like programming or entertainment. Previous studies have mainly focused on the potential downsides of Generative AI in educational settings, giving less attention to how it interacts with and is embraced by users in professional settings. This study fills this research gap by offering valuable insights for future studies.
Second, this study addresses a key issue in technology adoption research, which often concentrates solely on the advantages of new technologies. Models such as the TAM (Davis, 1989) and UTAUT (Williams et al., 2015) provide valuable perspectives on how the benefits of new technologies affect adoption decisions. However, these models consider only the facilitators of technology adoption, undermining the importance of inhibitors. To overcome this limitation, this study adopted the dual factor theory as a framework, examining both facilitators and inhibitors. The results indicate that the perception of a tangible threat plays a crucial role in discouraging users from embracing new technologies. This thorough investigation enhances our overall comprehension and lays a solid groundwork for future studies.

6.2. Practical implication

The study's findings suggest several practical implications for effectively encouraging the adoption of Generative AI in the workplace. Organizations should emphasize features related to knowledge support and entertainment. Specifically, promoting how Generative AI can aid in tasks like automating repetitive processes or providing instant information can enhance user acceptance. Additionally, highlighting the AI's capability to engage in interactive, entertaining communication can further drive its adoption. Companies should invest in showcasing these aspects through training and communication strategies.
The significant negative impact of realistic threat on adoption intentions indicates that employees might fear AI as a potential job disruptor. Organizations need to address these concerns proactively by clarifying how Generative AI can complement rather than replace human roles. Providing clear examples of AI's role in augmenting tasks rather than replacing them, and involving employees in the implementation process, can help mitigate these fears. Moreover, transparent communication about how AI will impact job roles and create new opportunities can alleviate concerns and foster a more positive reaction.

6.3. Limitations and Future Research

The study has pointed out some interesting discoveries, but there are certain limitations that suggest areas for future investigation. First, the sample was restricted to participants from the U.S. recruited through MTurk, which may not capture the full range of perspectives among global white-collar workers. To enhance the applicability of the findings, future studies should involve a more diverse and international sample. Moreover, this study considered white-collar workers from various industries and with differing levels of experience as a single, homogeneous group, which is a limitation. In addition, the skewed nature of the data regarding Generative AI tools and usage frequency is a limitation. Most participants used chat-based AI like ChatGPT, highlighting its current dominance in Generative AI. This suggests limited adoption of other AI types, such as image and audio generation in workplaces. Over half of respondents reported no AI usage, indicating many have yet to integrate these tools into their work routines. This imbalance may affect the generalizability of the findings across AI tools and usage patterns. Future research should use control or independent variables to explore differences across industries and sectors. A more balanced sample with a wider range of AI tools and use cases would provide clearer insights into Generative AI adoption.
Furthermore, while this research delved into factors like knowledge support, customization, entertainment value, perceived risk, realistic threat, and identity threat affecting Generative AI adoption, it overlooked other potential factors. To gain a more comprehensive understanding of adoption patterns, upcoming studies should explore additional drivers and barriers such as organizational culture, user experience aspects, and specific features of AI tools.
Lastly, this study focused on Generative AI adoption in a broad workplace setting. Future investigations could narrow down to specific industries or job roles to uncover the distinct challenges and opportunities presented by AI in diverse professional contexts. Examining industry-specific applications and their varied impacts on different job functions could yield more targeted insights and practical recommendations.

REFERENCES

Adiasto, K. 2024. SustAInable Employability: Sustainable Employability in the Age of Generative Artificial Intelligence. Group and Organization Management 10596011241238792 https://doi.org/10.1177/10596011241238792.
crossref
Adıgüzel, T., Kaya, MH., and Cansu, FK. 2023. Revolutionizing education with AI: Exploring the transformative potential of ChatGPT. Contemporary Educational Technology.

Aguinis, H., Villamor, I., and Ramani, RS. 2021. MTurk research: Review and recommendations. Journal of Management 47(4):823-837.
crossref pdf
Ashfaq, M., Yun, J., Yu, S., and Loureiro, SMC. 2020. I, Chatbot: Modeling the determinants of users’ satisfaction and continuance intention of AI-powered service agents. Telematics and Informatics 54: 101473 https://doi.org/10.1016/j.tele.2020.101473.
crossref
Ayinde, L., Wibowo, MP., Ravuri, B., and Emdad, FB. 2023. ChatGPT as an important tool in organizational management: A review of the literature. Business Information Review 40(3):137-149 https://doi.org/10.1177/02663821231187991.
crossref
Bianchi, C., and Andrews, L. 2018. Consumer engagement with retail firms through social media: an empirical study in Chile. International Journal of Retail & Distribution Management 46(4):364-385 https://doi.org/10.1108/IJRDM-02-2017-0035.
crossref
Bouhia, M., Rajaobelina, L., PromTep, S., Arcand, M., and Ricard, L. 2022. Drivers of privacy concerns when interacting with a chatbot in a customer service encounter. International Journal of Bank Marketing 40(6):1159-1181 https://doi.org/10.1108/IJBM-09-2021-0442.
crossref
Cao, J., and Song, Z. 2024. An incoming threat: the influence of automation potential on job insecurity. Asia-Pacific Journal of Business Administration ahead-of-print(ahead-of-print). https://doi.org/10.1108/APJBA-07-2022-0328.
crossref
Cenfetelli, Ronald T. 2004. Inhibitors and Enablers as Dual Factor Concepts in Technology Usage. Journal of the Association for Information Systems 5(11):https://doi.org/10.17705/1jais.00059.
crossref
Cheung, ML., Pires, G., Rosenberger III, PJ., Leung, WKS., and Chang, MK. 2021. The role of social media elements in driving co-creation and engagement. Asia Pacific Journal of Marketing and Logistics 33(10):1994-2018 https://doi.org/10.1108/APJML-03-2020-0176.
crossref
Cheung, ML, Pires, G, and Rosenberger, PJ 2020. The influence of perceived social media marketing elements on consumer-brand engagement and brand knowledge. Asia Pacific Journal of Marketing and Logistics 32(3):695-720 https://doi.org/10.1108/APJML-04-2019-0262.
crossref
Cohen, J 1988. Statistical power analysis for the behavioral sciences. 2nd edn. Erlbaum. Hillsdale.

Craig, K., Thatcher, JB., and Grover, V. 2019. The IT Identity Threat: A Conceptual Definition and Operational Measure. Journal of Management Information Systems 36(1):259-288 https://doi.org/10.1080/07421222.2018.1550561.
crossref
Dabbous, A, Aoun, Barakat K, and Merhej, Sayegh M 2022. Enabling organizational use of artificial intelligence: an employee perspective. Journal of Asia Business Studies 16(2):245-266 https://doi.org/10.1108/JABS-09-2020-0372.
crossref
Davis, FD 1989. Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology. MIS Quarterly 13(3):319-340 https://doi.org/10.2307/249008.
crossref
Dwivedi, YK, Kshetri, N, Hughes, L, Slade, EL, Jeyaraj, A, Kar, AK, Baabdullah, AM, Koohang, A, Raghavan, V, Ahuja, M, Albanna, H, Albashrawi, MA, Al-Busaidi, AS, Balakrishnan, J, Barlette, Y, Basu, S, Bose, I, Brooks, L, Buhalis, D, and Wright, R 2023. Opinion Paper: So what if ChatGPT wrote it? Multidisciplinary perspectives on opportunities, challenges and implications of generative conversational AI for research, practice and policy. International Journal of Information Management 71: 102642 https://doi.org/10.1016/j.ijinfomgt.2023.102642.
crossref
Fahad, SA, Salloum, SA, and Shaalan, K 2024. The Role of ChatGpt in Knowledge Sharing and Collaboration Within Digital Workplaces: A Systematic Review. In Al-Marzouqi A., Salloum S. A., Al-Saidat M., Aburayya A., Gupta B. (eds.), Artificial Intelligence in Education: The Power and Dangers of ChatGPT in the Classroom. pp 259-282. Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-52280-2_17.

Fornell, C., and Larcker, DF. 1981. Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research 18(1):39-50 https://doi.org/10.1177/002224378101800104.
crossref
Giordano, V., Spada, I., Chiarello, F., and Fantoni, G. 2024. The impact of ChatGPT on human skills: A quantitative study on twitter data. Technological Forecasting and Social Change 203: 123389 https://doi.org/10.1016/j.techfore.2024.123389.
crossref
Hair, JF, Risher, JJ, Sarstedt, M, and Ringle, CM 2019. When to use and how to report the results of PLS-SEM. European Business Review 31(1):2-24 https://doi.org/10.1108/EBR-11-2018-0203.
crossref
Halaweh, M 2023. ChatGPT in education: Strategies for responsible implementation.

Herzberg, F, Mausner, B, and Snyderman, BB 1959. The motivation to work. Wiley. New York, NY.

Huang, HL, Cheng, LK, Sun, PC, and Chou, SJ 2021. The Effects of Perceived Identity Threat and Realistic Threat on the Negative Attitudes and Usage Intentions Toward Hotel Service Robots: The Moderating Effect of the Robot’s Anthropomorphism. International Journal of Social Robotics 13(7):1599-1611 https://doi.org/10.1007/s12369-021-00752-2.
crossref
Hussain, K, Khan, ML, and Malik, A 2024. Exploring Audience Engagement with ChatGPT-Related Content on YouTube: Implications for Content Creators and AI Tool Developers. Digital Business 4(1):100071 https://doi.org/10.1016/j.digbus.2023.100071.
crossref
Isidore, C 2023. AI and actors strike: What it means for the entertainment industry. CNN; Retrieved from https://www.cnn.com/2023/07/18/business/ai-actors-strike/index.html.

Jing, Y, Wang, H, Chen, X, and Wang, C 2024. What factors will affect the effectiveness of using ChatGPT to solve programming problems? A quasi-experimental study. Humanities and Social Sciences Communications 11(1):319 https://doi.org/10.1057/s41599-024-02751-w.
crossref
Jo, H, and Park, DH 2023. AI in the Workplace: Examining the Effects of ChatGPT on Information Support and Knowledge Acquisition. International Journal of Human-Computer Interaction 1-16 https://doi.org/10.1080/10447318.2023.2278283.
crossref
Kasilingam, DL 2020. Understanding the attitude and intention to use smartphone chatbots for shopping. Technology in Society 62: 101280 https://doi.org/10.1016/j.techsoc.2020.101280.
crossref
Lee, MK, and Park, HJ 2019. Exploring Factors Influencing Usage Intention of Chatbot - Chatbot in Financial Service. Journal of Korean Society for Quality Management 47(4):755-765 https://doi.org/10.7469/JKSQM.2019.47.4.755.
crossref
Li, X, Hsieh, JJP-A, and Rai, A 2013. Motivational Differences Across Post-Acceptance Information System Usage Behaviors: An Investigation in the Business Intelligence Systems Context. Information Systems Research 24(3):659-682 https://doi.org/10.1287/isre.1120.0456.
crossref
Lim, WM, Gunasekara, A, Pallant, JL, Pallant, JI, and Pechenkina, E 2023. Generative AI and the future of education: Ragnarök or reformation? A paradoxical perspective from management educators. The International Journal of Management Education 21(2):100790 https://doi.org/10.1016/j.ijme.2023.100790.
crossref
Lin, HF 2011. An empirical investigation of mobile banking adoption: The effect of innovation attributes and knowledge-based trust. International Journal of Information Management 31(3):252-260 https://doi.org/10.1016/j.ijinfomgt.2010.07.006.
crossref
Lu, Y, Yang, S, Chau, PYK, and Cao, Y 2011. Dynamics between the trust transfer process and intention to use mobile payment services: A cross-environment perspective. Information & Management 48(8):393-403 https://doi.org/10.1016/j.im.2011.09.006.
crossref
Park, JS, and Park, HJ 2024. Understanding Post-Pandemic Travel Intention: Boredom as a Key Predictor. Journal of Korean Society for Quality Management 52(1):1-21 https://doi.org/10.7469/JKSQM.2024.52.1.1.
crossref
Park, JS, Yoo, JW, and Park, H 2024. Understanding user resistance of smart factory adoption: a focus on small and medium-sized enterprises. Asia Pacific Journal of Marketing and Logistics 36(7):1782-1800 https://doi.org/10.1108/APJML-09-2023-0896.
crossref
Park, MS, Park, JS, Yoo, JW, and Park, HJ 2022. Effects of the User Perception on Symbolic Adoption and Usage in Mandatory ATCIS-ǁ Use. Journal of Korean Society for Quality Management 50(3):517-532 https://doi.org/10.7469/JKSQM.2022.50.3.517.
crossref
Paul, J, Ueno, A, and Dennis, C 2023. ChatGPT and consumers: Benefits, Pitfalls and Future Research Agenda. International Journal of Consumer Studies 47(4):1213-1225 https://doi.org/10.1111/ijcs.12928.
crossref
Rudolph, J., Tan, S., and Tan, S. 2023. ChatGPT: Bullshit spewer or the end of traditional assessments in higher education? Journal of Applied Learning and Teaching 6(1):342-363.

Salvagno, M, Taccone, FS, and Gerli, AG 2023. Can artificial intelligence help for scientific writing? Critical Care 27(1):75 https://doi.org/10.1186/s13054-023-04380-2.
crossref pmid pmc
Shin, D 2021. Embodying algorithms, enactive artificial intelligence and the extended cognition: You can see as much as you know about algorithm. Journal of Information Science 49(1):18-31 https://doi.org/10.1177/0165551520985495.
crossref
Singh, S, Olson, ED, and Tsai, CH 2021. Use of service robots in an event setting: Understanding the role of social presence, eeriness, and identity threat. Journal of Hospitality and Tourism Management 49: 528-537 https://doi.org/10.1016/j.jhtm.2021.10.014.
crossref
Song, M, Du, J, Xing, X, and Mou, J 2022. Should the chatbot "save itself" or "be helped by others"? The influence of service recovery types on consumer perceptions of recovery satisfaction. Electronic Commerce Research and Applications 55: 101199 https://doi.org/10.1016/j.elerap.2022.101199.
crossref
Wang, X, Lin, X, and Shao, B 2023. Artificial intelligence changes the way we work: A close look at innovating with chatbots. Journal of the Association for Information Science and Technology 74(3):339-353 https://doi.org/10.1002/asi.24621.
crossref
Williams, MD, Rana, NP, and Dwivedi, YK 2015. The unified theory of acceptance and use of technology (UTAUT): a literature review. Journal of Enterprise Information Management 28(3):443-488 https://doi.org/10.1108/JEIM-09-2014-0088.
crossref
Wolf, V, and Maier, C 2024. ChatGPT usage in everyday life: A motivation-theoretic mixed-methods study. International Journal of Information Management 79: 102821 https://doi.org/10.1016/j.ijinfomgt.2024.102821.
crossref
Xiao, B, and Benbasat, I 2007. E-Commerce Product Recommendation Agents: Use, Characteristics, and Impact. MIS Quarterly 31(1):137-209 https://doi.org/10.2307/25148784.
crossref
Xu, Y, Zhou, G, Cai, R, and Gursoy, D 2024. When disclosing the artificial intelligence (AI) technology integration into service delivery backfires: Roles of fear of AI, identity threat and existential threat. International Journal of Hospitality Management 122: 103829 https://doi.org/10.1016/j.ijhm.2024.103829.
crossref
Yogeeswaran, K, Złotowski, J, Livingstone, M, Bartneck, C, Sumioka, H, and Ishiguro, H 2016. The interactive effects of robot anthropomorphism and robot ability on perceived threat and support for robotics research. Journal of Human-Robot Interaction 5(2):29-47 https://doi.org/10.5898/JHRI.5.2.Yogeeswaran.
crossref
Yoo, JW, and Park, HJ 2023. Understanding COVID-19 Vaccine Acceptance Intention: An Emotion-focused and Problem-focused Coping Perspective. Journal of Korean Society for Quality Management 51(4):643-662 https://doi.org/10.7469/JKSQM.2023.51.4.643.
crossref

Figure 1.
Proposed model framework
jksqm-52-3-509f1.jpg
Table 1.
Survey items
Variables Content Reference
Knowledge support SUP 1 : Using Generative AI tools like ChatGPT, I have access to more information than ever that is helpful for my job performance. Wang et al. (2023)
SUP 2 : Using Generative AI tools like ChatGPT, I have access to more knowledge than ever that is useful for my job performance.
SUP 3 : Through Generative AI, I have access to more resources than ever that help me enhance my job performance.
Customization CUS 1 : Generative AI tools enable customized services tailored to my needs in the workplace. Cheung et al. (2020)
CUS 2 : Generative AI tools provide lively and relevant information feeds that cater to my professional interests.
Entertainment ENT 1 : I find using ChatGPT to be enjoyable. Li et al. (2013); Wolf and Maier (2024)
ENT 2 : The actual process of using ChatGPT is pleasant.
ENT 3 : I have fun using ChatGPT.
Perceived risks RIS 1 : I do not feel totally safe providing personal private information over Generative AI chatbots. Lu et al. (2011)
RIS 2 : I am worried about using Generative AI chatbots because other people may be able to access my account
RIS 3 : I do not feel secure sending sensitive information through Generative AI chatbots.
RIS 4 : I do not feel safe providing work-related information over Generative AI chatbots.
Realistic threat REA 1 : The increased use of Generative AI in our everyday life is causing more job loss for humans. Huang et al. (2021)
REA 2 : In the long run, Generative AI poses a direct threat to human safety and well-being.
Identity threat IDT 1 : Recent advances in Generative AI technology are challenging the very essence of what it means to be human. Huang et al. (2021)
IDT 2 : Technological advancements in the area of Generative AI threaten human uniqueness.
Intention to use INT 1 : I am very likely to adopt Generative AI tools in the future. Lin (2011)
INT 2 : I plan to adopt Generative AI tools in the future.
INT 3 : I believe it is worthwhile for me to adopt Generative AI tools.
Table 2.
Demographic statistics of respondents (N=192)
Measure Value Frequency Percentage (%)
Gender Male 144 75
Female 48 25
Age group 20s 60 31.3
30s 84 43.7
40s 44 22.9
50s 3 1.6
60s 1 0.5
Education Lower than high school 7 3.7
High school graduate 128 66.7
College graduate 55 28.6
Master’s degree or above 2 1.0
AI policy AI types I do not want to use it, and it is prohibited at my workplace 3 1.6
I want to use it, but it is prohibited at my workplace. 87 45.3
I do not want to use it, but it is allowed at my workplace. 7 3.6
I want to use it, and it is allowed at my workplace. 23 12
I do not want to use it, but it is encouraged at my workplace. 1 0.5
I want to use it, and it is encouraged at my workplace. 71 37.0
Chat-based AI 154 80.2
Image generation AI 12 6.3
Audio generation AI 0 0.0
Video generation AI 0 0.0
Code generation AI 26 13.5
Usage frequency (per week) 0 times 99 51.6
1-2 times 41 21.4
3-4 times 28 14.6
5-6 times 10 5.2
Daily 14 7.3
Table 3.
Loadings and Convergent validity
Item Factor Loading Cronhach’s Alpha Composite Reliability AVE
Knowledge support SUP1 0.912 0.878 0.925 0.804
SUP2 0.862
SUP 3 0.916
Customization CUS1 0.915 0.777 0.899 0.817
CUS 2 0.893
Entertainment ENT1 0.933 0.897 0.936 0.830
ENT 2 0.882
ENT 3 0.917
Percevied risk RIS1 0.873 0.927 0.930 0.769
RIS2 0.772
RIS3 0.911
RIS4 0.941
Realistic threat REA1 0.959 0.827 0.915 0.844
REA2 0.876
Identity threat IDT1 0.963 0.915 0.924 0.859
IDT2 0.848
Usage intention INT1 0.921 0.934 0.958 0.883
INT2 0.947
INT3 0.950

Note : SUP: Knowledge Support, CUS : Customization, ENT : Entertainment, RIS: Perceived Risk, REA : Realistic threat, IDT : Identity Threat, INT : Intention to adopt Generative AI in workplace

Table 4.
Discriminant validity results (HTMT)
SUP CUS ENT RIS REA IDT INT
SUP
CUS 0.795
ENT 0.759 0.811
RIS 0.096 0.106 0.057
REA 0.137 0.258 0.236 0.204
IDT 0.146 0.15 0.076 0.751 0.151
INT 0.629 0.443 0.601 0.044 0.063 0.076

Note : SUP: Knowledge Support, CUS : Customization, ENT : Entertainment, RIS: Perceived Risk, REA : Realistic threat, IDT : Identity Threat, INT : Intention to adopt Generative AI in workplace

Table 5.
Hypotheses test results
β Confidence interval (2.5%, 97.5%) T-value P-value ƒ2 Hypothesis support
H1: Knowledge support → Usage intention 0.411 [0.027, 0.719] 2.348 0.019 0.137 Supported
H2: Customization → Usage intention -0.110 [-0.432, 0.230] 0.655 0.513 0.010 Not supported
H3: Entertainment → Usage intention 0.400 [-0.008, 0.702] 2.220 0.026 0.124 Supported
H4: Perceived risk → Usage intention 0.028 [-0.201, 0.175] 0.281 0.778 0.001 Not supported
H5: Realistic threat → Usage intention -0.187 [-0.308, 0.008] 2.345 0.019 0.055 Supported
H6: Identity threat → Usage intention 0.013 [-0.154, 0.170] 0.154 0.878 0.001 Not Supported
TOOLS
PDF Links  PDF Links
PubReader  PubReader
ePub Link  ePub Link
Full text via DOI  Full text via DOI
Download Citation  Download Citation
  Print
Share:      
METRICS
0
Crossref
632
View
23
Download
Related article
Editorial Office
1806, 310, Gangnam-daero, Gangnam-gu, Seoul, 06253, Korea
TEL: +82-2-563-0357   FAX: +82-2-563-0358   E-mail: ksqmeditor@ksqm.org
About |  Browse Articles |  Current Issue |  For Authors and Reviewers
Copyright © The Korean Society for Quality Management.                 Developed in M2PI
Close layer
prev next