Data Justice in Practice: A Guide for Developers

The Advancing Data Justice Research and Practice project aims to broaden understanding of the social, historical, cultural, political, and economic forces that contribute to discrimination and inequity in contemporary ecologies of data collection, governance, and use. This is the consultation draft of a guide for developers and organisations, which are producing, procuring, or using data-intensive technologies. It provides actionable information for those who wish to implement the principles and priorities of data justice in their data practices and within their data innovation ecosystems. In the first section, we introduce the nascent field of data justice, from its early discussions to more recent proposals to relocate understandings of what data justice means. This section includes an account of the outreach we conducted with stakeholders throughout the world in developing a nuanced and pluralistic conception of data justice and concludes with a description of the six pillars of data justice around which this guidance revolves. Next, to support developers in designing, developing, and deploying responsible and equitable data-intensive and AI/ML systems, we outline the AI/ML project lifecycle through a sociotechnical lens, walking the reader through each phase and noting the ethics and governance considerations that should occur at each step of the way. This portion of the guide is intended to provide a background picture of the different stages of the lifecycle and to show how the data justice pillars can be woven into the stages and their respective sociotechnical considerations. To support the operationalisation data justice throughout the entirety of the AI/ML lifecycle and within data innovation ecosystems, we then present five overarching principles of responsible, equitable, and trustworthy data research and innovation practices, the SAFE-D principles—Safety, Accountability, Fairness, Explainability, and Data Quality, Integrity, Protection, and Privacy. These principles support and underwrite the advancement of data justice within research and innovation practices. We elaborate upon them as high-level goals that are then followed by further specification through the presentation of additional properties, which are to be established in either the project or the system to ensure these goals are reached. Depending on their contexts, potential impacts, and scale, data innovation activities should be carried out in a way that involves different degrees of stakeholder engagement. To facilitate this process, the next section provides an explainer of the Stakeholder Engagement Process and the steps it includes—preliminary horizon scanning, project scoping and stakeholder analysis, positionality reflection, and establishing stakeholder engagement objectives and methods. Finally, the last section presents guiding questions that will help developers both address data justice issues throughout the AI/ML lifecycle and engage in reflective innovation practices that ensure the design, development, and deployment of responsible and equitable data-intensive and AI/ML systems. This is done by presenting questions related to both the six pillars of data justice and the SAFE-D principles introduced previously.This report was commissioned by the International Centre of Expertise in Montréal in collaboration with GPAI's Data Governance Working Group, and produced by the Alan Turing Institute. The research was supported, in part, by a grant from ESRC (ES/T007354/1), Towards Turing 2.0 under the EPSRC Grant EP/W037211/1, and from the public funds that make the Turing's Public Policy Programme possible