Abstract
Digital phenotyping, or personal sensing, is a field of research that seeks to quantify traits and characteristics of people using digital technologies, usually for health care purposes. In this commentary, we discuss emerging ethical issues regarding the use of social media as training data for artificial intelligence (AI) models used for digital phenotyping. In particular, we describe the ethical need for explicit consent from social media users, particularly in cases where sensitive information such as labels related to neurodiversity are scraped. We also advocate for the use of community-based participatory design principles when developing health care AI models using social media data.
Reference10 articles.
1. Back to the Future: Achieving Health Equity Through Health Informatics and Digital Health
2. Towards Informed Practice in HCI for Development
3. The growing field of digital psychiatry: current evidence and the future of apps, social media, chatbots, and virtual reality
4. Notice of Special Interest (NOSI): Computational and statistical methods to enhance discovery from health data (NOT-LM-23-001)National Institutes of Health2024-04-01https://grants.nih.gov/grants/guide/notice-files/NOT-LM-23-001.html
5. Notice of Special Interest (NOSI): Addressing health disparities in NIMHD research: leveraging health data science (NOT-OD-22-026)National Institutes of Health2024-04-01https://grants.nih.gov/grants/guide/notice-files/not-od-22-026.html