보안 품질과 효율 품질이 성과에 미치는 영향

Abstract

Abstract

Artificial intelligence (AI), particularly generative AI, has emerged as a core technology driving digital innovation and operational efficiency across industries, due to its ability to generate human-like content. However, the widespread adoption of this technology also introduces new security threats. Generative AI operates by learning from vast amounts of data, which means that personal information or corporate secrets may be collected and stored through prompt inputs. This increases the risk of unintended data leakage or security breaches such as information exposure or data corruption. This study aims to identify the factors that influence user performance based on security quality, efficiency quality, and security enhancement among members of the general public who use or have an interest in generative AI. Based on prior research, the components of security quality were categorized as confidentiality, integrity, non-repudiation, accountability, and authenticity. Efficiency quality was defined by time behaviour, resource utilization, and capacity. The model also included security enhancement and user performance as outcome variables. The survey targeted adults over the age of 20 who have experience with or understanding of generative AI. A total of 390 online questionnaires were distributed, with 352 valid responses used for final analysis. Structural equation modeling was conducted using Smart PLS 4.0. The findings indicate that: First, the hypothesis that security quality significantly affects security enhancement was partially supported. confidentiality, integrity, non-repudiation, and accountability had statistically significant impacts on security enhancement, whereas authenticity did not show a statistically significant influence. This indicates that, despite its technical importance, users perceive authenticity as not being directly related to security enhancement. Second, the effect of efficiency quality on security enhancement was also partially supported. Time behaviour and resource utilization exhibited statistically significant positive impacts, while capacity did not have a statistically significant impact. This implies that time behaviour and resource utilization are more closely linked to security enhancement than system capacity. Third, the impact of security enhancement on user performance was fully supported. This indicates that security enhancement contributes to improved overall work efficiency, which in turn leads to enhanced user performance. The implications of this study are as follows: Generative AI is rapidly spreading across industries due to its ability to automatically generate content. However, this widespread adoption also entails risks such as data breaches and increased security vulnerabilities. To maximize user performance in the use of generative AI, it is essential to enhance security capabilities while recognizing both efficiency quality and security quality as equally important factors.

Keywords: Generative AI, Security Quality, Efficiency Quality, Security Enhancement, User Performance.