Predictive analytics in HIV surveillance require new approaches to data ethics, rights, and regulation in public health.

In recent years, applications of big data-driven predictive analytics in public health programs have expanded, offering promises of greater efficiency and improved outcomes. This commentary considers the turn toward predictive modeling in US-based HIV public health initiatives. Through two case studies, we analyze emergent ethical problems and risks. We focus on potential harms related to (1) classifying people living with HIV in public health systems, (2) new ways of combining and sharing individuals' health data that predictive approaches employ, and (3) how new applications of big data in public health challenge the underlying logics and regulatory paradigms that govern data re-uses and rights in public health practice. Drawing on critical technology scholarship, critical bioethics, and advocacy by organized networks of people living with HIV, we argue that stakeholders should enter into a new range of reform-oriented conversations about the regulatory frameworks, ethical norms, and best practices that govern re-uses of HIV public health data in the era of predictive public health interventions that target individuals. [ABSTRACT FROM AUTHOR]