AI 활용 관련 표준에 대한 개인정보보호 관점에서의 적정성 검토

The advancement of AI technology enables AI to learn from data and generate new information on its own. As a result, the use of AI is increasing due to its positive effects, such as improving work efficiency, protecting data and privacy, and creating economic value. However, generative AI not only raises economic and ethical concerns but also poses risks of privacy infringement, and negative use cases such as actual privacy violations are on the rise. To address these issues, organizations such as NIST, the EU, the OECD, and the Personal Information Protection Commission have established standards for the safe use of AI. However, due to differences in the scope and purpose of these standards, as well as varying national perspectives on AI, it is difficult to identify the appropriate standard for an organization from a personal data protection standpoint. This study analyzes domestic and international standards based on criteria such as the application of PbD, inclusion of privacy-related keywords, and the presence of AI utilization standards and methodologies. It identifies the characteristics of each standard and pinpoints those that can serve as references from a privacy perspective. Furthermore, it aims to identify any shortcomings in the identified standards and propose improvement measures. It is expected that utilizing the findings of this study to consider both efficiency and privacy protection when applying AI technology will contribute to sustainable AI development.