Continuance use of AI coding assistants among South Korean Industry Developers: A survey case study with large language models

Utilizing generative artificial intelligence including large language models (LLMs), into programming tasks has recently garnered significant attention. This paper investigates the motivations determining the adoption of LLM-oriented services, integrating two user-oriented concepts, the technology acceptance model and expectation confirmation theory into the foundational frameworks. By combining these models, the study presents a comprehensive framework for understanding developers’ continual usage intentions of these technologies. Analyzing data from 1,338 developers, the study reveals that developers are more inclined to adopt LLM-oriented services for their programming tasks. Key findings include the critical role of perceived ease of use, the alignment of service performance with initial expectations, and the influence of developers’ enjoyable feeling on their acceptance. These insights emphasize the importance of ensuring that LLM-oriented services not only meet developers’ expectations but also enhance their overall programming experience through ease of use and enjoyment.