AI 시대 디지털 거버넌스에의 헌법주의 재구성

In a phase where Artificial Intelligence (AI) is restructuring the information order, the realm of expression, and market structures, traditional constitutionalism faces criticism for failing to adequately capture the issue of the “publicization of private power.” As traditional constitutional principles struggle to fully protect individual rights in the AI era, “Digital Constitutionalism” emerges as a crucial alternative. It is defined as a theoretical and practical framework that projects the core values of modern constitutionalism—fundamental rights, the rule of law, and democratic legitimacy—onto the specificities of digital society. Furthermore, it can be characterized as a project of multi-layered constitutionalization applicable across state, trans-state, and private spheres. The international community has long embedded digital constitutionalism within digital norms. Since the OECD AI Principles(2019) and the UNESCO Recommendation on the Ethics of AI(2021) established a global common denominator centered on human rights, transparency, safety, and accountability, the European Union(EU) has asserted itself as a global standard-setter. Through the Digital Services Act(DSA), the EU established procedural rights and transparency as general obligations for platforms and heightened duties for Very Large Online Platforms(VLOPs), thereby seeking the substantive realization of user rights. Furthermore, the EU AI Act introduced risk-based regulation and a phased application of transparency and governance obligations for general-purpose AI. Influenced largely by EU law, South Korea overhauled its national governance by implementing the Framework Act on Artificial Intelligence in January 2026, which imposes impact assessments and risk management obligations on high-impact AI, as well as labeling requirements for generative AI. Additionally, the amendment to the Act on Promotion of Information and Communications Network Utilization and Information Protection, effective July 2026, introduced regulations on disinformation and hate speech (illegal information), strengthening provider accountability through obligations for reporting, measures, and transparency reports for large-scale platforms. While both Korean statutes share the digital constitutionalist goals of controlling private power and guaranteeing fundamental rights, they face critical issues due to escalating tensions regarding regulatory layers (risk vs. content), procedural control, and proportionality. Most notably, delegating legislative content to private autonomous policies and imposing excessive liability without clear criteria for restricted subjects makes it difficult for these laws to remain free from controversies regarding unconstitutionality. Such recent legislation, which bypasses direct regulation of AI and platforms through “private regulation,” ultimately results in state authority evading the Constitution and the rule of law. This leads to the infringement of fundamental rights, a chilling effect on freedom of expression in digital spaces, and the rise of surveillance capitalism. This study is rooted in this critical awareness. It emphasizes the necessity of digital constitutionalism in domestic AI and platform regulation and examines its applicability. For South Korea's AI and platform legal system to substantively incorporate elements of digital constitutionalism, efforts must be made to introduce basic principles of rights protection—such as the guarantee of freedom of expression, self-determination, and due process—as overarching concepts even within “service norms,” such as terms of use and service policies executed in the private sphere. Furthermore, specific core elements applicable to the digital space must be standardized among the regulated entities.