Skip to yearly menu bar Skip to main content


Poster
in
Workshop: Regulatable ML: Towards Bridging the Gaps between Machine Learning Research and Regulations

Multilingual Compliance: A Comparative Study of Privacy Policies in Chinese, Japanese, and Korean

Muhammad Hassan · Phillip Nakamura · Yuanye Ma


Abstract:

As global data privacy regulations tighten, privacy policies have become pivotal in shaping the relationship between users and online services. Despite their importance, these documents remain inaccessible to many users due to excessive length and complex legal terminology, posing significant challenges to informed consent. Moreover, most existing research has concentrated on English-language policies, leaving a gap in our understanding of how privacy policies are structured and enforced in non-English-speaking regions.To address this, we present a study that examines privacy policies from Chinese, Japanese, and Korean websites, analyzing their compliance with respective national privacy laws. We collected and processed a dataset of over 2,400 privacy policies, and more than 5.2 million tokens, employing language detection and text analysis techniques to assess adherence to regulatory frameworks, including China’s PIPL, Japan’s APPI, and Korea’s PIPA. Our findings reveal notable disparities in regulatory compliance, with Chinese and Japanese policies generally demonstrating stronger alignment with legal standards compared to Korean policies.These results highlight the need for enhanced clarity and enforcement across regions, offering insights into the evolving global landscape of data privacy. Our work underscores the importance of multilingual analysis in advancing the accessibility and transparency of privacy practices, laying the groundwork for future research and policy development in this critical domain.

Chat is not available.