AI 서비스 관리 요인과 AI 서비스 보호 요인의 중요도 우선순위 분석

Artificial intelligence (AI) is developing into a core technology that will change the paradigm of our time following the Internet and mobile revolution. From the interactive AI service to the Generative AI era, its use in all fields of society is rapidly spreading. While AI technology has the potential to reorganize the entire industry, there are also unreliable threats of AI services. AI-based attacks are becoming more sophisticated and easy, and there can also be cases where AI is exploited to confuse society, such as malicious codes and phishing sites using AI. Therefore, it is also necessary to approach the management and protection of measures to detect security threats using AI in real time and reduce the development and risk of responding technology. In this study, we intend to derive important factors by approaching it in order to safely utilize AI in consideration of the information protection and personal information protection management system that AI service companies should observe. It is intended to derive the priorities of important factors for items for AI service management factors and AI service protection factors. AI service management factors were selected as security policy establishment factors and risk management system establishment factors, and AI service protection factors were selected as AI education and training, AI system access control, learning data quality and security secured, automated result verification and reliability, AI system log and monitoring, and AI system backup and recovery measures. The research tool used for empirical analysis was a questionnaire, and the survey subjects consisted of field security experts who are in charge of developing, evaluating, and managing security technologies by security-related companies and institutions. The questionnaire was distributed to a total of 20 experts, and the period of distribution and collection of the questionnaire was from mid-May to mid-June 2025, and the final 20 copies were collected. In addition, for research analysis, an empirical analysis was performed by applying the AHP analysis technique that evaluates hierarchical analytical decision-making. Based on the results of this study, important priorities for AI service management and protection activities can be identified, and it is expected that this will reflect important factors that are effective in changing policy systems for AI service management and protection activities.