Discrimination of Algorithmic Decision-Making and Protection of Human Rights : With a focus on the legal regulation of Europe

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.