Ethics surrounding open data include numerous stakeholders, from clinicians, individual researchers generating data, researchers accessing shared data, data curators, universities/institutes, governing bodies, and the individuals the data are collected from. On top of these entities, there is an additional layer of geography that brings in cultural/societal influences about what data and with whom data can be shared. The multiplicity of perspectives cannot be adequately presented in a panel, and in such a varied landscape, the egalitarian format of a round table would encourage the diversity of shared perspectives. It is important to get a diverse array of perspectives because while the contexts in which we create sharable data are different, the data serves a common purpose to help further our understanding of the human brain. It is understandable that researchers look at this complex landscape and conclude the safe ethical course of action is to keep the data local and not share it for fear of privacy/legal challenges or other potential consequences. However, not sharing data also has the ethical consequence of inhibiting scientific progress.

The technology to share data has vastly outpaced the conversation around what should be shared, with whom, and how. How can we share data that: maximizes the usefulness of that data, offers due credit to the data generators and curators, and respects the privacy and rights of those we’ve collected the data from? Large consortiums have collected vast amounts of data and are in flux on how to distribute the data to aid research. In addition to the large consortiums of data, there is a long tail of smaller datasets generated by individual labs that are just as important but do not have the resources to spend on navigating the minutia of legal/ethical barriers. Some platforms can host this long tail of data, but there is hesitation due to the aforementioned barriers. Sharing data is not an all-or-nothing choice, and even if data is shared, that does not give researchers carte blanche to run any analysis on the data.

As in many dilemmas, there is no free lunch, but in a round table discussion, there is an opportunity to navigate toward shared solutions.There will be a couple of key concepts/terms that will need to be defined to have a productive discussion: anonymized data, de-identified data, federated analysis, and a high-level awareness of governmental laws protecting data rights.

OSR 2025 | Table Talk 3: Ethical Issues in Open Science: How Open should Open Data be?