Investigating the Cross-app Data Sharing Implications of Using Android Health Connect
Mobile health and fitness (mHealth) apps are widely used, and they process sensitive user data. To improve the interoperability among them and enable fine-grained permission controls for users, Google strategically introduced an on-device centralized mHealth data sharing platform, Android Health Connect (AHC), in 2022. However, we argue that the current AHC design may exacerbate privacy risks because it enables broad but implicit mHealth data sharing across apps. In this article, we first conduct a cross-app data sharing study on 41 AHC-compatible apps; by analyzing their declared AHC-related permissions and privacy practice disclosures, we reveal that they can broadly and intensively share different types of mHealth data with each other but do not adequately disclose AHC-related practices. We further conduct an online user study for 105 participants to interact with our simulated AHC permission management interfaces for two apps; by quantitatively and qualitatively analyzing participants’ interactions and their responses to the follow-up questions, we demonstrate that most participants have a flawed understanding of AHC’s permission model and its implicit cross-app data sharing implications. Our findings highlight the pressing need to improve the design of this new and pivotal AHC platform to better protect mHealth app users’ privacy.
Authors
- Anhao Xiang (ORCID: https://orcid.org/0000-0002-8850-8583)
- Weiping Pei (ORCID: https://orcid.org/0000-0002-6831-2313)
- Chuan Yue (ORCID: https://orcid.org/0000-0002-6095-4768)
- Madelyn Swelstad (ORCID: https://orcid.org/0009-0000-1892-7523)
- Andrew Plute (ORCID: https://orcid.org/0009-0000-5711-8526)
Institutions
- Colorado School of Mines (US)
- Kennesaw State University (US)
- University of Tulsa (US)
Publication Details
- Journal
- ACM Transactions on Computing for Healthcare
- Published
- 2026-07-23
- DOI
- https://doi.org/10.1145/3831593
- Primary Topic
- Mobile Health and mHealth Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00