Case study 4 : Online radicalisation: detecting weak signals before indoctrination
The context: online radicalisation
Yanis, 16, is going through a difficult period after changing secondary school. Quiet and passionate about current affairs, he spends more and more time on forums and video platforms. His father, Mourad, notices harsher, more divisive talk, but thinks it is a teenage phase.
The problem: gradual ideological influence
It is not a sudden shift, but repeated exposure to polarising content:
Algorithmic echo chamber
Successive recommendations of increasingly radical content.
Victimising rhetoric
An “us versus them” narrative.
Normalisation of violence
Gradual justification of extreme views.
The wake-up call
One evening at a family dinner, Yanis says worrying things about “the need to act”. Troubled, Mourad checks the Vigiloo app and finds a moderate-level alert flagging repeated exposure to highly ideological content.
How Vigiloo helped
The system identified:
Analysis of content viewed
Abnormal frequency of videos and forums with extreme rhetoric.
Change in vocabulary
Increasing use of polarising terms in online exchanges.
Preventive alert
Notification encouraging dialogue before any break.
The results
A calm conversation between Yanis and his father, without confrontation.
More balanced recommendation settings.
A gradual return to more nuanced exchanges and social ties maintained.
