상세 보기
인공지능 시스템 신뢰성 감리 프레임워크
- 정찬우;
- 장항배;
- 주영지;
- 김정덕
초록
As artificial intelligence (AI) technology rapidly develops and expands, its societal and economic effects also increase. However, issues of AI misuse, social dysfunction, and deepfakes have arisen, highlighting the need for developing trustworthy AI systems. While ethical standards for trustworthy AI have been proposed by major global institutions, the development of specific and applicable AI reliability audit criteria from a system perspective remains highly inadequate. This study aims to establish the concept of AI reliability and develop a framework and checklist for AI system development projects. Through a literature review, we identified five key attributes of AI reliability and proposed an AI system reliability audit framework consisting of three perspectives that can evaluate them. Furthermore, we developed an AI system reliability audit checklist based on the framework and evaluated its importance and applicability using the FGI analysis.
키워드
- 제목
- 인공지능 시스템 신뢰성 감리 프레임워크
- 제목 (타언어)
- A Study of Artificial Intelligence System Reliability Audit Framework
- 저자
- 정찬우; 장항배; 주영지; 김정덕
- 발행일
- 2023-04
- 저널명
- 한국산업보안연구
- 권
- 13
- 호
- 1
- 페이지
- 29 ~ 54
- 언어
- KOR
- 출판사
- 한국산업보안연구학회
- 발행국가
- 대한민국
- 분량
- 26 페이지
- ISSN
- P 2765-2327