- 英文摘要
- Insider trading has long been recognized as one of the most serious violations of securities market regulations because it undermines market integrity, damages investor confidence, and weakens the fairness and efficiency of capital markets. Despite the importance of maintaining consistency in criminal sanctions, sentencing outcomes in insider trading cases often vary considerably due to differences in judicial discretion, case complexity, and the absence of standardized sentencing criteria. In response to these challenges, this study examines the sentencing factors applied in insider trading cases in Taiwan and undertakes a comparative legal analysis of insider trading regulatory frameworks and sentencing practices in the United States, the United Kingdom, and Germany using artificial intelligence (AI)-assisted analytical methods. Based on these findings, the study proposes recommendations for developing a robust, transparent, and adaptive AI-assisted sentencing model for insider trading cases.
Given that the Judicial Yuan of Taiwan has not yet established formal sentencing guidelines specifically for insider trading offenses, this study adopts empirical methods to identify key sentencing factors relevant to AI- assisted analysis and evaluate their influence on sentencing outcomes. Through the collection and examination of judicial decisions, the study seeks to establish an evidence-based framework that supports more consistent sentencing practices while reducing potential bias and promoting fairness, proportionality, and legal certainty.
The development and training of an AI-assisted sentencing model for insider trading cases may provide substantial benefits to judicial practice. Such a system could assist judges in handling increasingly complex financial crime cases involving large volumes of evidence, sophisticated financial transactions, and specialized legal and economic knowledge. By providing data-driven sentencing recommendations and analytical support, AI-assisted tools may improve judicial efficiency while preserving judicial discretion and accountability.
Because insider trading cases frequently involve complex assessments of illicit profits, market impact, and subjective intent, there is an urgent need to explore the application of modern AI technologies to criminal sentencing. An effective AI-assisted sentencing system has the potential to enhance the accuracy and consistency of judicial decisions, reduce judicial workload and administrative burdens, improve court operations, and ultimately strengthen public confidence in the administration of justice.
- 目次
- 壹、緒論
一、研究背景
二、研究動機
貳、內線交易規範量刑原則
一、內線交易規範
二、刑罰目的
三、AI 人工智慧量刑系統原則
參、外國內線交易量刑法制研究
一、國際內線交易法制
二、外國量刑法制分析
肆、我國內線交易罪之量刑因子探討與 AI 系統建構分析
一、Artificial Intelligence 量刑分析結果
二、本國內線交易罪影響量刑的因素
伍、結論與建議
