Introduction

This high-level guidance explores ethical considerations associated with the use of machine learning techniques for research and statistical purposes.  This guidance is not exhaustive, but aims to assist and support analysts, researchers, data scientists, and statisticians navigating the ethical issues surrounding machine learning based projects. Links to further resources are provided if you would like to read about particular aspects in more detail.

This guidance has been produced in collaboration with the United Nations Economic Commission for Europe (UNECE) Machine Learning 2021 group led by the United Kingdom’s Data Science Campus, and as part of their Data Ethics workstream. This workstream was led by the UK Statistics Authority’s data ethics team and involved engagement and collaboration with multiple countries. This workstream was developed following the identification of a need for further applied guidance in the use of machine learning for the production of official statistics by the international research and statistical community. The guidance has been created with a focus on the use of machine learning tools for the undertaking of research and production of aggregate statistics (recommendations are made at a policy level, not an individual level) and whilst there may be some overlap with operational or other uses, this is not the intended focus of the document.

Though many of the ethical topics discussed within this document are not unique to machine learning, they are nonetheless worth further consideration with a focus on the practical application of mitigations specific to machine learning. The guidance is divided into several parts, providing an initial introduction to machine learning and ethics, the most common ethical considerations needing exploration when using machine learning, some potential mitigations for these issues, and links to further resources.

An ethics checklist has also been provided, which summarises the main points covered in this guidance.