본문 바로가기 주메뉴 바로가기
국회도서관 홈으로 정보검색 소장정보 검색

결과 내 검색

동의어 포함

목차보기

Title page 1

Contents 5

Foreword 4

Executive summary 7

1. Introduction to RBC due diligence and key considerations for AI 8

Introduction to responsible AI 9

Purpose of this guidance 9

Target audience 10

Understanding the risks related to the development and use of AI 13

Characteristics of trustworthy AI 14

Basics of RBC due diligence 14

How to use this guidance 18

2. Due diligence framework and practical examples for identifying and addressing risks 19

Step 1 - Embed RBC into policies and management systems 20

Step 2 - Identify and assess actual and potential adverse impacts 24

Step 3 - Cease, prevent and mitigate adverse impacts 34

Step 4 - Track implementation and results of due diligence activities 47

Step 5 - Communicate actions to address impacts 48

Step 6 - Provide for or co-operate in remediation when appropriate 49

References 51

Glossary 54

Notes 58

Tables 5

Table 2.1. Step 1: Roadmap of related provisions in existing frameworks 20

Table 2.2. Step 2: Roadmap of related provisions in existing frameworks 24

Table 2.3. Factors to consider when prioritising risk 33

Table 2.4. Step 3: Roadmap of related provisions in existing frameworks 34

Table 2.5. Step 4: Roadmap of related provisions in existing frameworks 47

Table 2.6. Step 5: Roadmap of related provisions in existing frameworks 48

Table 2.7. Step 6: Roadmap of related provisions in existing frameworks 49

Figures 5

Figure 1.1. Graphical representation of the RBC due diligence framework 15

Figure 2.1. Due diligence expectations based on involvement with the adverse impact 31

Boxes 6

Box 1.1. Considerations for Small and Medium Sized Enterprises (SMEs) 12

Box 2.1. Examples of high-risk uses of AI systems drawn from different frameworks 26

Box 2.2. Understanding the risk-based approach 27

Box 2.3. Identifying risks to data quality, interoperability and access throughout the AI system lifecycle 28

Box 2.4. Understanding involvement with the risk 30

Box 2.5. Scenarios illustrating the involvement framework 32

Box 2.6. Tailoring risk management to the enterprise's circumstances 34

Box 2.7. Using AI to support RBC due diligence 35

Box 2.8. Enabling transparency, explainability and traceability throughout the AI system lifecycle 37

Box 2.9. Content authentication and provenance mechanisms 38

Box 2.10. Pre-deployment response plan 39

Box 2.11. Preventing or mitigating risks when deploying AI systems 40

Box 2.12. Deployment in contexts where laws are inconsistent with international standards on RBC 41

Box 2.13. Temporary or permanent suspension of the functioning of the AI system 41

Box 2.14. Special considerations for enterprises engaging with 'control points' 44

Box 2.15. Understanding disengagement from business relationships in the context of risks 45

Box 2.16. Practical examples of due diligence for investors and financial institutions investing in the development of AI systems 46

Box 2.17. Potential options for remedying adverse impacts 50

출판사 책소개

알라딘제공
This report provides practical guidance to enterprises for implementing OECD standards on responsible business conduct (RBC) and the OECD AI Principles when developing and using artificial intelligence (AI). It aims to support innovation, investment and growth of enterprises in the AI value chain by helping enterprises proactively address adverse impacts. The report promotes policy coherence, and where possible interoperability, between the OECD and other national or international AI risk management frameworks.