去中心化虛擬資產服務開發者刑事責任之研究——以洗錢防制法為中心

Abstract

本文以具非託管架構、自動執行之智慧合約及非集中式中介者等技術特徵之去中心化虛擬資產服務為探討對象。當此類技術遭使用於掩飾犯罪所得時，開發者是否屬洗錢防制法之義務主體，抑或可能成立洗錢罪之正犯或幫助犯，乃待釐清之核心問題。 循此問題意識，爰選取美國Helix案、荷蘭 Tornado Cash（Pertsev 案）及美國 Tornado Cash（Storm 案）一審判決進行分析。Helix 案涉及開發者實際經營混幣服務，屬典型中心化營運模式；Tornado Cash 案則涉及智慧合約部署後由使用者自行操作，開發者已無直接控制資金之權限，其是否應因系統設計、治理或維護行為而負擔刑事責任，係論證之爭點所在。 技術開發是否即構成洗錢犯罪，本文認為仍應回歸技術開發者於資金處理機制中所處之功能地位，及其對資金流向所之實質影響力，依個案具體事實加以判斷。不宜僅因使用者嗣後利用該技術從事洗錢犯罪即逕認開發者應負刑事責任。 換言之，唯有開發者進行之技術或制度設計，對資金移轉、隱匿、分層或整合等洗錢行為具有足以左右犯罪結果實現之支配力，並透過權限配置、系統架構設計或治理機制之安排，使該支配力具體展現於洗錢行為之完成，且因設計或持續優化而使洗錢犯罪構成要件之實現具有實質促進或強化作用而直接或間接獲取利益，始具備刑事歸責之基礎。 同時，亦應進一步檢視開發者與洗錢犯罪之可罰違法性，始能在「技術開發」與「犯罪參與／犯罪支配」間劃定合理界線，使刑事責任之歸屬建立於開發者之實際功能角色與行為貢獻，而非僅依技術形式或抽象風險加以推定。

This article focuses on a particular category of decentralized virtual asset services—one characterized by non-custodial architecture, self-executing smart contracts, and the absence of centralized intermediaries. When such technologies are exploited to conceal criminal proceeds, a central question arises as to whether their developers qualify as obligated entities under anti-money laundering law, or may instead be held liable as principals or accessories to the crime of money laundering—this is the core issue this article seeks to clarify. To address this issue, the article analyzes the first-instance decisions in United States v. Harmon (the Helix case), the Dutch Tornado Cash case (Pertsev), and the U.S. Tornado Cash case (Storm). The Helix case involved a developer who actually operated a mixing service, representing a typical centralized operational model. The Tornado Cash cases, by contrast, involved smart contracts that, once deployed, were operated independently by users, such that the developers no longer possessed direct control over the funds. Whether such developers may nonetheless bear criminal liability on account of their system design, governance, or maintenance activities constitutes the focal point of this article's inquiry. This article contends that whether a technology developer commits the crime of money laundering should still be determined, on a case-by-case basis, by reference to the functional role the developer plays within the fund-processing mechanism and the substantive influence the developer exercises over the flow of funds; it would be inappropriate to hold a developer criminally liable merely because a user subsequently exploits the technology to launder money. In other words, a basis for criminal attribution exists only where the developer's technical or institutional design possesses a degree of control sufficient to determine the realization of the criminal result at the various stages of money laundering—namely, transfer, concealment, layering, and integration—of funds; where, through the arrangement of permission configurations, system architecture, or governance mechanisms, that control is concretely manifested in the completion of the money laundering conduct; and where, as a result of the design or its continued optimization, the developer substantially facilitates or reinforces the fulfillment of the elements of the money laundering offense, thereby directly or indirectly obtaining a benefit. At the same time, further examination should be given to the degree of connection between the developer's conduct and the realization of the punishable unlawfulness of money laundering, as well as to the concrete grounds for attributing culpability to the developer, so as to draw a reasonable line between "technology development" and "participation in crime" or "control over the crime." This would ensure that the attribution of criminal liability is grounded in the developer's actual functional role and behavioral contribution, rather than being presumed merely on the basis of the technology's formal characteristics or an abstract level of risk.