Data Ethics

Data ethics is the set of standards and working practices that guide how data is collected, stored, analysed, shared, and used, with attention to the people it describes and affects. It applies across the whole lifecycle, not only at collection, and it asks questions that the law may not: whether a use is appropriate, who benefits, who bears the risk, and whether the people involved would recognize what is being done.

Why personal judgment is not enough

Analysts bring assumptions and blind spots to their work, and a decision that feels obviously fine to one team can look very different to the people described in the data. Ethics becomes checkable when it is written down as standards, applied through review by someone outside the team, and recorded with the reasons for each decision. The UK government’s Data Ethics Framework, written for public-sector projects, organises this around transparency, accountability, and fairness, and asks teams to define the benefit, understand the applicable law, use no more data than the purpose needs, be open about data use, and evaluate what happened.

For a fictional bus operator, the ethical questions about reusing fare-card taps for crowding analysis are not answered by “it is legal” or “it is useful.” They are answered by writing the purpose, the minimum fields, riders’ reasonable expectations, the effect on drivers, the alternative of collecting nothing new, and the name of the person who decides.

Distinct from law, privacy, and fairness testing

Legal compliance is a floor; a use can be lawful and still fail an organization’s ethical standard, and the law differs by jurisdiction, so an ethics plan records which law applies rather than assuming one. Data privacy is one of the concerns data ethics covers, alongside transparency about commercial use, fairness of outcomes, and the responsibilities of those who hold data. Testing a model for bias is a specific technical activity that an ethics standard may require; it is not the whole of data ethics.

Claims such as “people own their raw data” or “open data is ethical data” are slogans rather than standards. Replace them with the specific rights, duties, and checks that apply to the case at hand.

Reference: UK Government: Data Ethics Framework (2020). Examples here are illustrative.