Jessica Pidoux: “Dating apps are not magic. Just maths, code and sensitive data.”
Jessica Pidoux is a doctoral student in digital humanities at the École Polytechnique Fédérale de Lausanne, where she wrote her thesis on online dating services. At the same time, she set up the data collective Dating Privacy. Interview with this sociologist-geek-humanist whose mission is to match the interests of dating applications with those of their users.
by Charles Foucault-Dumas published on 02 June 2021
What is Dating Privacy?
It's a collective motivated by the will to understand and improve users’ personal data management by dating applications. To conduct my PhD research entitled “Online dating quantification practices. A human-machine learning process”, I interviewed 40 users of 26 different dating applications. A variety of issues were raised in relation to privacy: geolocation tracking when the app was disabled, exposure to undesirable user profiles, accounts blocked without any justification, and, more broadly, ignorance about how the apps present results. Some users never obtained a match or went on a date. These issues do not have yet a clear answer to users.
The personal data collection and processing by these applications remains opaque. This is a difficulty to assert users’ rights. Users have the right to know what information feeds the system. That is why, quite spontaneously, users and myself got together to first map the personal data collected by these dating app services and understand how it is used (including for hidden purposes beyond dating, like advertisement).
Today, there is a discrepancy between what dating apps claim in their privacy policies or in the media and the personal data they actually collect through the self-description forms (which are much more extensive than for any other digital service). Amongst all the sensitive information provided, there is no visibility about the criteria used to present one user to another (hair colour or income level). By law, the user has the right to know if s-he is the subject of automated decisions and how these are made.
What does Dating Privacy actually do to solve these problems?
Dating Privacy is addressing four goals:
- Mapping the data collected by dating applications
- Raising awareness among a maximum number of users by making this mapping as visible as possible through our own publications and articles in the media
- Conceptualising the notion of privacy and thus be able to define users' rights in the specific context of dating applications
- Providing the dating app user community with the tools to assert these rights and take back power over their data
All these efforts will contribute to privacy and algorithmic literacy in the context of dating apps. We tackle in particular how users are tracked via geolocation, what it means to obtain dating app results that are biased or to be the subject of an algorithmic discrimination in a dating app.
What does “taking back power over their data” mean in practice?
The current truth is that companies behind dating applications dispossess us of our personal data. Accessing it is a very complex process. You must send a written request invoking the GDPR law. They respond (when they respond) with a file that can be incomplete or unreadable for a non-technical person. And if you are able to understand it, you realize that they sent you what data these companies have on you but not what they do with it. In other words, data is no longer ours.
Taking back the power over our data means having a tool at our disposal not only to see and understand what information these applications have about us, but also to choose what we want to share and for what purposes. Some dating apps like Tinder now have dedicated pages to request our personal data. But it is impossible to know if everything that is collected is really provided to us and how the app uses this information. Our collective’s researches show that some apps are not consistent with the information they actually collect and what they say to users. Consequently, data protection and privacy should be independently designed, with users’ own control in mind and not just the company’s legal compliance. Dating Privacy will be that intermediary.
Thanks to the Dating Privacy community, users also share their data with other users to improve their understanding about the app they use. This is another goal of our collective: the exchange of experience, good practices, knowledge... a real common place, unlike dating apps that are designed to retain users in the app to maximise their profit. These apps are “match” machines playing on their opacity to maintain a kind of belief in their magic. Except that there is no magic, online dating is a quantifiable practice that relies on sensitive data to function.
Who is concerned?
Everyone who uses a dating application is concerned and invited to join us. But we are also working to make the tools created for Dating Privacy usable for issues encountered in other platforms. Amazon’s or Netflix’s recommendation algorithms, for example, have many similarities in with those introducing one person to another. Again, no magic. Just maths, code and data. The impact of the work done by Dating Privacy will thus go beyond dating applications alone.
You invite dating app users to join you. How can they participate?
With its members from several European countries, our community is active online: we have a forum, a Wiki, a Twitter account, a fledgling blog and, soon, a newsletter. Joining this conversation is a first way to participate. A second way is to request your personal data through our request generator. We can help you analyse the file you receive, without retrieving your data: we are developing tools and methods to empower you to do it. These practical cases will then feed the collective.
Curious to know what dating apps really know about you? Ask for your data now.
All good wills are also welcome: individuals who want to participate in the tools development or to push further the ones we have already created, but also those who defend the same causes as us. I think of the NGOs which try to protect communities that these applications make (more) vulnerable. Dating Privacy can provide them with the means to assert their rights.
What is the benefit for dating application users to share their data with Dating Privacy?
This is a subtle issue. People download an app to find a partner or dates, not to lift the lid and study the mechanics. Nor to recover their data and give it to someone else. However, it is important to make people aware of the risks involved in having a public profile on these applications: phishing, identity theft via all the data entered (profile picture, job, education, age, sex...), harassment, discrimination... It is also important to know that this data feeds the advertising ecosystem and therefore influences our online experiences, including browsing outside the dating applications. Without sharing our “data facts” is not possible to find the roots of the problem.
Dating Privacy also tackles very practical user problems: “I don't have a match”, “I never get a message?” Analysing, together with the user, the data obtained from the platform of their choice, means understanding why, whether the answers are to be found in the design of the application or in the use made of it. For example, Tinder is competitive: the most active members are favoured, they appear more often in the results list.
What is the role of HestiaLabs in your collective?
With its expertise in privacy and its technical skills, HestiaLabs is a support structure for Dating Privacy. HestiaLabs has already developed subject access request (SAR) generators. These tools simplify and automate the process of data reappropriation by users, giving them the means to defend their interests. The aim of this collaboration is also that HestiaLabs can reuse the developments and knowledge created by Dating Privacy to help other data collectives. HestiaLabs also has an international network of experts from different backgrounds that help us answer legal and engineering questions.
Interview by Charles Foucault-Dumas.