Responsible Data Governance of Neuroscience Big Data

Abstract

Current discussions of the ethical aspects of big data are shaped by concerns regarding the social consequences of both the widespread adoption of machine learning and the ways in which biases in data can be replicated and perpetuated. We instead focus here on the ethical issues arising from the use of big data in international neuroscience collaborations. Neuroscience innovation relies upon neuroinformatics, large-scale data collection and analysis enabled by novel and emergent technologies. Each step of this work involves aspects of ethics, ranging from concerns for adherence to informed consent or animal protection principles and issues of data re-use at the stage of data collection, to data protection and privacy during data processing and analysis, and issues of attribution and intellectual property at the data-sharing and publication stages. Significant dilemmas and challenges with far-reaching implications are also inherent, including reconciling the ethical imperative for openness and validation with data protection compliance and considering future innovation trajectories or the potential for misuse of research results. Furthermore, these issues are subject to local interpretations within different ethical cultures applying diverse legal systems emphasising different aspects. Neuroscience big data require a concerted approach to research across boundaries, wherein ethical aspects are integrated within a transparent, dialogical data governance process. We address this by developing the concept of “responsible data governance,” applying the principles of Responsible Research and Innovation (RRI) to the challenges presented by the governance of neuroscience big data in the Human Brain Project (HBP).

From Responsible Research and Innovation to responsibility by design

“…To illustrate, the HBP is, among other things, a big data integration and provision project (Fothergill et al 2019a). Data governance is therefore key to the project and to the development of the infrastructure.…”

“…The HBP's final phase started in April 2020 and will end in 2023. We have provided detailed accounts of our work during the first 8 years elsewhere Aicardi et al 2018;Fothergill et al 2019a;Stahl et al 2019;Aicardi et al 2020;Salles and Farisco 2020). We have published Opinions on three topics: data protection and privacy (Salles et al 2021), responsible dual use (Aicardi, Bitsch, Bådum, et al 2021), and artificial intelligence (AI) (Aicardi, Bitsch, Burton, et al 2021).…”

Research and Practice of AI Ethics: A Case Study Approach Juxtaposing Academic Discourse with Organisational Reality

“…It is worth noting that a large-scale survey in Australia in 2020 indicated that 57.9% of end-users are not at all confident that most companies take adequate steps to protect user data. The significance of understanding and employing responsibility is an issue targeted in many studies (Chatfield et al, 2017;Fothergill et al, 2019;Jirotka et al, 2017;Pellé & Reber, 2015). Trust and control over BD + AI as an issue is reiterated by a recent ICO report demonstrating that most UK citizens do not trust organisations with their data (ICO, 2017).…”

Addressing Ethical Issues in AI

“…In the health field, for example, where the sensitivity of patient and health data is universally acknowledged, data management and data governance are explicitly seen as ways of ensuring ethical goals (Rosenbaum 2010, OECD 2017. The proximity of ethical concerns and data governance has also led to the development of data governance approaches that are explicitly developed around ethical premises (Fothergill et al 2019).…”