인공지능시대 알고리즘 의사결정에 의한 차별과 인권보호방안에 관한 연구 — 유럽의 법적 규제에 관한 논의를 중심으로 — = Discrimination of Algorithmic Decision-Making and Protection of Human Rights — With a focus on the legal regulation of Europe —

다국어 초록 (Multilingual Abstract)

In the information age, advances in artificial intelligence technology are bringing increased efficiency and convenience to our daily lives. The utilization of automated algorithmic decision-making has become a common phenomenon in both the public and private sectors. However, problems arising by algorithms of artificial intelligence have also been in various forms, and have been caused due to the transparency and fairness of algorithms.
This paper analyzes the provisions of anti-discrimination law and data protection law as the most appropriate legal instruments in Europe to protect unlawful discrimination to algorithmic decision-making. While anti-discrimination law can be used as an effective regulatory tool to address not only intentional discrimination but also indirect discrimination by algorithms, it is limited by the difficulty in objectively proving the legitimacy of discrimination due to the non-transparency of algorithms. In addition, while Data protection laws can also mitigate the risk of unfair and unlawful discrimination by requiring openness and transparency in the use of personal data, limitation of data protection laws include questions of applicability and the practical impact of data protection regulations on automated decision-making. While proper enforcement of anti-discrimination laws and data protection laws can protect people from discrimination by algorithms, it is important to note the limitations of these laws as a means of regulating algorithmic decision-making.
In this regard, the European Union has enacted the AI Act, a comprehensive regulatory law to protect human rights, including equality and non-discrimination as stipulated in the EU Treaties, and to protect fundamental human rights, democracy, the rule of law, and environmental sustainability from high-risk AI. The AI Act enhances transparency on how algorithms are operated from the design stage of AI models and mitigates the shortcomings of anti-discrimination laws in terms of difficulties in proving discrimination as a result of algorithms and remedies for the consequences of AI decision-making. Even if the AI Act will be phased in as a comprehensive regulatory tool for AI technology, AI systems such as Chat GPT will continue to be subject to a period of grace, and the issue of balancing fairness and privacy during the training and evaluation phase of AI, as well as the appropriateness of fairness metrics in relation to biases that may lead to discrimination, will continue to be a focus of attention.

참고문헌 (Reference)

• 1 김건우, "차별에서 공정으로: 인공지능의 차별완화와 공정성 제고를 위한 제도적 방안" 전북대학교 법학연구소 2019
• 2 오정하 ; 김일환, "지능정보사회에서 알고리즘의 법･규제의 방향" 청주대학교 법과 정책 연구소 27 (27): 2021
• 3 한국연구재단, "주요 국가의 인공지능관련 연구윤리 정책 동향조사" 2020
• 4 양종모, "인공지능의 윤리/정책/사회 이슈" KAIST 2019
• 5 김효은, "인공지능과 윤리" 박영사 2024
• 6 손도일, "알고리즘 통제: 투명성의 구체적 기준, DAIG 제2호, 2021신옥주, 유럽인권협약에 의한 기본권 보장" 2009
• 7 홍성욱, "새로운 정보기술과 인권" 한국과학기술한림원 2018
• 8 Sandra Wachter, "Why fairness cannot be automated: Bridging the gap between eu non-discrimination law and ai" 41 : 2021
• 9 Sandra Wachter, "Why a right to explanation of automated decision-making does not exist in the General Data Protection Regulation" 7 (7): 2017
• 10 Frederik J. Zuiderveen Borgesius, "The international journal of Human Rights" Routledge 2020
• 11 Philipp. Hacker, "Teaching fairness to artificial intelligence: Existing and novel strategies against algorithmic discrimination under EU law" 55 : 2018
• 12 Paul De Hert, "Reinventing data protection?" Springer Science 2009
• 13 "REGULATION (EU) 2016/679 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation)"
• 14 Aaron Rieke, "Public Scrutiny of Automated Decisions: Early Lessons and Emersing Methods, An Upturm and Omidyar Network Report"
• 15 "Proposal for a Regulation laying down harmonised rules on artificial intelligence (Artificial Inteligence Act)"
• 16 Walter L. Perry, "Predictive Policing: The Role of Crime Forecasting in Law Enforcement Operations" RAND’s publications 2013
• 17 Indr ˙e Zliobait ˙e, "Measuring discrimination in algorithmic decision making" University of Helsinki
• 18 J. Angwin, "Machine Bias: There’s Software Used Across the Country to Predict Future Criminals. And It’s Biased Against Blacks"
• 19 Luca Deck, "Implications of the AI Act for Non-Discrimination Law and Algorithmic Fairness"
• 20 "European Parliament 2019-2024 Texts Adopted P9_TA(2024)0138 Artificial intelligence Act, Whereas (1), (2)"
• 21 함인선, "EU 데이터보호판례" 마로니에 2023
• 22 Sandra Fredman, "Discrimination Law" Oxford University Press 2011
• 23 M. Kusner, "Counterfactual Fairness, Proceedings of the 31st International Conference on Neural Information Processing Systems"
• 24 "Convention 108+, Convention for the protection of individuals with regard to processing of personal data"
• 25 H. Weerts, "Algorithmic unfairness through the lens of eu non-discrimination law: Or why the law is not a decision tree" 2023
• 26 이병규, "AI의 예측능력과 재범알고리즘의 헌법적 문제-State v. Loomis판결을 중심으로" 21 (21): 2020
• 27 "AI Act: a step closer to the first rules on AI"