Evaluating Third-Party Involvement in Android Apps: Norms and Anomalies in Usage Patterns

This paper tackles the pressing challenges in the digital realm, with a specific focus on privacy and security concerns in Android apps, especially regarding third-party involvement and data usage disclosures. A key issue addressed is the complexity and inaccessibility of privacy policies, which often remain obscure to users due to their technical legal language. To address these challenges, we employ cutting-edge artificial intelligence techniques, notably Generative AI (GenAI) and natural language processing (NLP), to decipher and interpret the privacy policies of Android apps with a focus on third-party data practices. A central feature of our study is the introduction of an innovative classification system that segregates app behaviors into ‘Transparent’ and ‘Opaque’ categories. This distinction is rooted in the clarity and explicitness of third-party data practice disclosures within app privacy policies. ‘Transparent’ apps are those that clearly articulate their interactions with third-party entities and the purposes for data usage, aligning with standard practices and enhancing user trust. Conversely, ‘Opaque’ apps lack clear disclosures, leaving users uninformed about potential data sharing and usage, thereby raising privacy and security concerns. Our research contributes by utilizing the capabilities of ChatGPT’s API for an in-depth analysis of privacy policies, with a particular emphasis on third-party mentions and data usage purposes.