Comparative Analysis on The Level of Vagueness in Instagram and Tiktok Privacy Policies

Abstract

Social media platforms handle vast amounts of user data, yet privacy policies that govern this data practices are often shrouded in vagueness. This study investigated vagueness embedded within the privacy policies of two prominent social media platforms, specifically Instagram and TikTok, aiming to empower users with greater awareness and understanding regarding privacy. By employing a quantitative approach with content analysis method, this study seeks to shed light on the prevalence and the level of vagueness within the two privacy policies. Drawing upon semantics theory, this research identified vague terms within each policy based on Bhatia et al.'s taxonomy then measured the vagueness level according to the scoring system developed by Reidenberg et al. The analysis revealed a significant presence of vagueness, stemming from Modality terms (e.g., "may", "can"), Numeric Quantifiers (e.g., "some", "certain"), Conditional terms (e.g., "applicable"), and Generalization terms (e.g., "general"). Both platforms found heavily relied on Modality terms, creating uncertainty around data practices permissions and possibilities. Furthermore, the analysis yielded a Vagueness Score of 1,212 for Instagram and 1,311 for TikTok, implying a slightly higher level of vagueness in TikTok privacy policy. Interestingly, despite containing more vague terms, Instagram privacy policy had a lower Vagueness Score due to the presence of a larger proportion of clear statements within the longer document. Overall, the research highlights the critical issue of vague language in social media privacy policies, hindering user comprehension. The findings emphasize the need for clear and transparent communication between platforms and users.