Foucault in the Cafe: The Mobile Phone and Consent to Surveillance

On a Friday afternoon, in a cafe on Dorcol, I sat at a table by the window. I had a coffee in front of me and a small notebook. At the next table, nine people were sitting around two joined wooden tables. Each of them, at some point during the forty minutes I spent there, looked at a phone. It was not an isolated scene; it was the normal rhythm of the cafe.

Out of professional habit, I tried to count how many times each of them picked up the phone. By a conservative count, the average number of phone pick-ups per person over forty minutes was twenty-one. The highest was a woman of about thirty-five, who picked up the phone thirty-four times. The lowest was a man of about sixty, who picked it up only nine times but never once put it down off the table.

What struck me in the scene was not the count. What struck me was that each of them was, at that moment, actively connected to a system of apps, while none of them could have said which app was actually running, which data were being transmitted, to whom and for what purpose.

A pivot to Foucault

I pivot to the sociology of surveillance as formulated by the French philosopher Michel Foucault. In the book “Discipline and Punish: The Birth of the Prison” (Gallimard, 1975; Serbian translation, Karpos, 2016), Foucault developed the concept of the panopticon, borrowed from Jeremy Bentham’s prison design of 1791. Foucault’s central claim runs roughly as follows:

What matters is that the person being watched knows they may be watched at any moment, but can never be certain they are being watched at that particular moment. That uncertainty is enough to reshape behaviour, even when no watcher is actively present.

Foucault wrote about the panopticon as an architecture of the prison, the factory, the school, the hospital. He wrote in the seventies, when there was as yet no internet, no mobile phone, no app. But in the digital age his analysis has become more precise than he could ever have predicted.

The nine people at the table on Dorcol are not under active surveillance. No one is watching them. There is no central operator reading their messages. But they behave as if they could be watched. Nobody writes a political message they do not want someone to see. Nobody types a sentence they do not want an algorithm to keep. Nobody searches for content they do not want recorded in a history of browsing.

Consent was given, formally, when each app was installed. But the consent is not remembered, because it was packed into a twenty-page document that nobody reads.

What apps actually know

To avoid abstraction, I will list what a typical Serbian app today processes while it sits in the background of your phone. The list is drawn from reading the terms of service of four major Serbian digital brands, which I will not name individually, because this analysis is not about them but about the structure:

• precise location updated every few seconds
• the list of other apps open in the same period
• a purchase history from other apps with the same owner
• the time you wake and the time you fall asleep, inferred from the day’s first and last taps
• the network of contacts you exchange messages with, ranked by frequency

This is not a theoretical framework. This is standard functionality. Consent is given on first launch, usually with a single tap, usually without reading the text.

The counter-argument about choice

Here comes the counter-argument one hears often when the subject of digital surveillance is raised. By that logic, the user is free not to install the app. If the terms are unpleasant, the choice exists not to use it. Nobody is being forced onto social networks, taxi apps or food-delivery apps.

The argument is formally accurate, but sociologically it is not honest. Life in an urban Serbian environment in 2026 without at least five basic apps is, in practice, a form of isolation that people cannot afford. A banking app is almost compulsory, because the city centre no longer has enough teller windows. A food-delivery app is an option neighbours and colleagues use, so refusal marks a social difference, not just a practical one. A group-chat app is where family communication happens, so opting out means losing touch with relatives.

Foucault wrote about this only obliquely. Consent is not free when refusing consent means losing a basic social function. That is, as he put it, the discipline of consent, which operates without violence yet operates.

Our silences about our own transparency

I return to the nine people at the table on Dorcol. At one point the woman of about thirty-five showed something on her phone to her partner beside her. He looked, laughed, and went back to his coffee. I wondered whether he would have laughed in the same way had he known that the content had at that moment been transmitted to a server of some company in Frankfurt, where it became part of a model predicting what he will buy in a year’s time, how he will vote in the next election, and whether his marriage is approaching a crisis.

I am not writing this to provoke paranoia. I am writing it to set out a sociological finding. Serbian society, like many others, is walking through the largest change in the structure of privacy in history with complete calm, as if nothing were happening.

The question that remains, which I share with those who have read this far, runs roughly like this. If consent has not been honestly given because refusal is in practice impossible, is it consent at all. And, the harder sociological question, how many of us would be willing if, say, tomorrow that consent began to be strictly enforced, with every message readable on a public screen before we sent it.