인권보장을 위한 인공지능 입법방향 - 디지털 권리장전에 비춰본 영향분석의 필요성 -

Attempts to establish new AI normative frameworks are being made around the world. While it is true that there are trials of direct AI regulatory lawmaking in some cases, it can be said that the mainstream is to present normative standards and principles for AI-based society. In the end, these are likely to be referred as specific directions for AI-related legislation, which is expected to become more serious in the future. Therefore, at this stage, it is necessary to concretely establish the purpose of the AI normative frameworks. Ultimately, the purpose of such normative frameworks is to ensure substantive human rights. In this context, South Korea recently promulgated the Digital Bill of Rights. However, by focusing only on the instrumental role of human rights, it merely rephrased the rights already guaranteed through international norms and constitutions. In order to contribute to legislation in the AI era that can actually guarantee human rights, it is necessary to proactively propose a process to analyze or assess the impact and build a value consensus based on it, rather than simply relying on normative declarations. In particular, the need to institutionalize impact assessment cannot be overemphasized in that the context of AI technology and its use inevitably contains indeterminacy, and requires a risk-based approach rather than standardized legal regulation.