When we talk about privacy problems, we tend to focus on the collection or use of personal data without that person’s informed consent – for instance, hacking, unauthorised access, staff (mis)behaviour, or automated data collection.
However, a person’s privacy may be compromised even when they willingly agreed to share their data. Dan Nunan and MariaLaura Di Domenico discussed two such situations, in a paper published in the International Journal of Market Research (available here). One situation is when individual datasets are combined; the other the emergence of new analytical techniques.
Threat one: Combining data sets
An individual data set may not present a privacy problem on its own, but could become problematic when combined with other sets. Nunan and Domenico described a specific case where researchers matched anonymous information on a dating website with public Facebook photos, by using facial recognition software. This allowed them to obtain significant insight about those persons, again from information publicly available on Facebook. So, while a person may be satisfied with the level of personal information that they provided on the dating website, and with the level of personal information that they provided on Facebook, their position may be different when the two data sets were combined.
Threat two: New analytical techniques
A data set may not present a privacy problem in the present moment, but could become problematic in the future, as analytical techniques keep developing and offering new ways of look at data. One example could be Instagram posts. At the moment, many companies do not have the ability to analyse images at a large scale and, so, most users may feel comfortable with the content that they are sharing on that social media platform. However, such content will remain available online, and will be searchable, for the foreseeable future. As technology evolves rapidly, and companies develop new capabilities, soon it will be feasible for many companies to process these data in a cost-effective manner.
In my view, data shared ABOUT children on social media (the so called sharenting phenomenon) suffers from these two threats: there is a lot of very personal and very revealing information being shared by parents, relatives, friends, schools… And when these children become adults, and consumers in their own rights, they will have inherited a very detailed online representation, which they will find very difficult to erase – much like a tattoo… only, they did not get to choose this one.
What will be the impact for these children of the accumulation of such publicly available information about them? For instance, when these children become adults, will they be at a disadvantage when trying to get health insurance because it is public information that they wet their beds until quite late?
Difficult questions which have been on my mind, lately, and for which I do not have an answer…
Ana Canhoto 