HDI goes beyond the short-sighted focus on ‘transparency’ of data sharing/analytics — a concept that side-steps difficult questions such as whether people can really understand a complex system made transparent — and instead addresses subsequent choices, actions and effects. It is intended to bring ethics to the forefront of systems design, and to frame traditional ethical concerns—such as autonomy, consent and agency—in ways that speak directly to the technical community.
There is a lot of rhetoric about ethical frameworks for the design of such systems, but very little of this has been practically oriented, i.e. giving clear help to designers considering how to design a system so as to make its data collection, processing and sharing understandable and useful to those whose data is collected, processed and shared. HDI is such a practically oriented framework, and centres on three tenets; legibility, agency and negotiability.
Legibility – Making the processes of sharing data about a person, and others’ analysis and use of that data, comprehensible to that person.
Agency – Giving a person the capacity to interact with their systems so as to control and correct the above-mentioned processes;
Negotiability – Giving a person the capacity to interact with the people who do the above-mentioned analysis and use, so as to change and correct what those people do.
Please note the use of these three words varies in different areas, but we have a specific focus here. A February 2020 survey paper on HDI, by Victorelli et al., reinforces the view of Mortier and others, that “HDI refers to the analysis of the individual and collective decisions we make and the actions we take, whether as users of online systems or as subjects of data collection”, and that the term implies an explicit link between individuals and the signals they emit.
For more on the conceptual framework of HDI, set out around 2013 in papers by Mortier and others, see here and here.