Skip to content
Log in
Sponsored Have an interesting project or opportunity? → Hire me ↗
Sponsored Design Multi-Agent AI Systems Using MCP and A2A: Engineer your own Pyt Design Multi-Agent AI Systems Using MCP and A2A: Engineer your own Pyt

Author

Sihyun Kim

View on Semantic Scholar

2 papers indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Open access Jul 2026

Defining High‐Impact AI in Autonomous Shipping: Legal Frameworks, Supply Chain Resilience, and Maritime Safety Dimensions

The rapid emergence of Maritime Autonomous Surface Ships (MASS) has made the governance of artificial intelligence systems a question of immediate regulatory urgency. The European Union's Artificial Intelligence Act (2024) and the Republic of Korea's AI Framework Act (2025) are the first two horizontal, binding, risk‐tiered AI statutes in force, yet neither provides operationally adequate criteria for classifying AI systems aboard autonomous vessels. Maritime transport sits outside the EU Act's Annex III, and Korea's Act defers the operative classification to guidelines whose maritime treatment remains incomplete. This paper proposes five interrelated criteria for classifying autonomous shipping AI as high‐impact—functional centrality to navigation and safety, reliability under maritime stress, systemic harm potential upon malfunction, dependence on real‐time data integrity, and structural absence of human override—grounded in the operational realities of autonomous vessels and the legal obligations both frameworks activate once the threshold is crossed. Because roughly four‐fifths of world trade by volume moves by sea through a small number of congested chokepoints, accurate classification is also a question of supply chain resilience, with dimensions extending beyond technical risk to seafarer welfare, marine‐environmental integrity, and the democratic legitimacy of a transition whose pace has outrun the governance architecture intended to steer it.

Sihyun Kim, Tae Jung Park · 1 citation
Open access Jul 2026

The shared blind spot: why diverse AI governance approaches fail for the same reason

AI governance instruments are proliferating, and so are their difficulties. Across major jurisdictions and international bodies, reform efforts built on substantially different premises encounter a recognizably similar pattern of failure. I argue that anticipatory regulatory governance rests on three operational premises —categorical stability, epistemic accessibility, and manageable pace—and that AI’s emergence, opacity, and velocity violate all three in compound. These premises form a distinct layer of operational preconditions, not a complete theory of governance. Reform within the existing premises reproduces the violations they produce. I call this configuration the reform trap : a paradigmatic lock-in at the level of operational preconditions, distinct from path dependence and policy paradigm rigidity. The pattern is convergent across five strategies in active reform—categorical regulation, process-based management, information disclosure, normative guidance, and adaptive experimentation—and persists even in the most adaptive of them. I propose three premise-level substitutions —outcome observability, causal attributability, and enforcement capability—each replacing a violated precondition with a weaker, design-addressable one. These differ from outcome-based and performance-based regulation, which swaps instruments within an architecture whose premises remain stable.

Sihyun Kim · 0 citations