The archive · Product Ideas · Product decision · 2025
Movistar's Game Guardian Discord bot catches adults faking children's voices in games
Movistar built a free Discord bot that detects AI-cloned child voices and alerts moderators and parents in real time, instead of just warning them.
Movistar
What it had to solve
Artificial intelligence lets adults clone children's voices, so predators can join gaming voice chats sounding like kids — practically undetectable by parents or moderators. Movistar Argentina, the internet provider many young gamers use, wanted to protect children rather than just broadcast warnings about the new threat.
How it works
Voice-cloning AI gave online groomers a new disguise: an adult can now sound like a child inside gaming voice chats, beyond what parents or moderators can detect by ear. Movistar Argentina and VML Argentina chose not to make another warning film, but to build the protection itself — a free tool that hears what humans cannot.
Game Guardian works as a Discord bot with three parallel systems: detection of voices generated by the leading voice-emulation platforms, a language model trained on grooming patterns such as asking age, location or for photos, and continuous transcription of the conversation. When the systems agree the voice is fake, moderators are alerted; after three consecutive danger signals, parents automatically receive a full transcript by email, even if they are not in the room.
The launch confronted parents directly: they heard what sounded like a child, then learned it was an AI clone. Early operations on live servers logged more than 4,500 protected players, 72 automated alerts and 14 active cases escalated to moderators. Five months after launch the tool had won over 50 industry awards, including three Golds and four Silvers at El Ojo de Iberoamérica 2025.
Why it lands
- A functional bot gave families something to do, not just something to fear.
- Training detection on real voice-cloning platforms and grooming language made the tool specific rather than symbolic.
- Alerting moderators and emailing parents transcripts created a protection network around each child.
- The reveal experience turned an abstract AI threat into something parents felt in a minute.
What it did
In its first months on live Discord servers, the tool protected more than 4,500 players, issued 72 automated alerts, enabled moderator action on 14 active cases and registered 453 emergency contacts. Within five months of launch it had gathered more than 50 industry awards, including three Golds and four Silvers at El Ojo de Iberoamérica 2025 and a Grand Prix at the Diente awards.
What you can take
A brand can solve a fear it cannot talk away: by turning AI from the predator's tool into a parent's tool, Movistar made child safety something a community installs rather than a message it watches.
Since then
Game Guardian stayed free and available worldwide on Discord, and Movistar Argentina framed it as proof of its role as guardian of safe connectivity for the young gamers it serves. VML's teams described it as the first voice-focused barrier against AI-assisted grooming, and the tool's award run — capped by three Golds and four Silvers at El Ojo de Iberoamérica 2025 — made it one of the most decorated Latin American ideas of the year.
Sources
- Game Guardian – Case Study
- El éxito de 'Game Guardian', la herramienta de IA creada por VML Argentina y Movistar para combatir el grooming
spotted an error? The archive wants to know.
Your turn
You just read one. Describe the brief you are staring at, and see who has been given the same problem.
Free account · 3 free questions · no card