Surveillance of social networks: the state misses TikTok, and tells us why

According to franceinfo, which has published an investigation by "L'Œil du 20H" (a current-affairs strand of the French evening news), police are struggling to pick out the most viral TikTok videos during a wave of social anger. The reason, according to the channel: a change of social-media monitoring software. The new tool is cheaper, but it would also be less effective. Social networks, franceinfo reminds us, serve as a fast means of communication for protesters, but also for some "casseurs" (hooligans or rioters who smash things up at demonstrations). The title of the investigation quotes a harsh verdict ("C'est une hérésie", roughly "It's heresy"), but the summary we have does not say who said it. We also don't know the name of the software, its price, or exactly what it misses. Everything that follows is our reading.

First, the framing. The problem, in this story, is that the watching machine isn't watching well enough. Nobody asks whether the machine should be watching at all. This is common sense at work: surveillance is presented as a technical given, and the debate shrinks to the quality of the kit. The reasonable citizen, in this account, is the one who worries that the police can't see clearly.

Yet social anger did not wait for the software to exist. The videos circulating on TikTok are, for those who film and share them, a way of telling what is happening in their own words: a small factory of narrative, outside the newsrooms. For Gramsci, hegemony is not just a matter of the truncheon; it is the capacity to make one's own view of the world pass for normality. A movement that produces its own images, live, nibbles away at that monopoly. When a state wants to spot what goes viral, it also wants to know which story is winning.

Is the police looking for offences or for narratives? Both, no doubt, and the line is blurred. The investigation mentions the casseurs, a convenient figure that justifies the whole apparatus; what it does not say, at least in the summary, is what happens to the monitoring of all the other videos, those of ordinary protesters. The tool, for its part, does not sort by good intentions.

Then there is the most telling detail: the economics. A cheaper, less effective piece of software is austerity applied to surveillance. You almost feel sorry for the administration, condemned to managing its own gaze like an accountant. But irony has its limits: if the tool improves, it will be with public money, and without any debate about what it is for.

The dominant ideas belong to those who also control the means of producing meaning. Here, the platform belongs to private interests, the monitoring tool to a contractor, and the story of the protest circulates between the two. The real question is therefore not "why can the police see so badly?" but "who decides what deserves to be seen, and in the name of which order?"