Our Theater of Anonymity

Abstract

Ethics review boards are increasingly asked to review big health data research proposals using a regulatory framework written prior to the current era of machine learning and artificial intelligence. Traditional consideration of individual identifiability does not account for the growing recognition that almost all data can be reidentified. This leaves the research ethics community performing a “theater of anonymity”: weighing benefit versus risk on the inclusion of participant identifiers alone. The wider research community, including U.S. federal agencies, is pushing for greater transparency and data sharing, stretching the current definition of identifiability to the breaking point. Additionally, shifting attitudes on the balance between privacy and research benefit may make a project more acceptable to individuals and communities than to the ethics boards designed to serve them. Reviewing the human research community's historical and current understanding of identity in terms of research, risk, benefit, and consent may point toward a need for a change in how consent is conceived. Particularly in the context of consent for big health data research, the research ethics community may need to shift its focus from solely considering individuals toward actively considering the interests of the communities most likely to be affected by the research.