Whenever a teenager misbehaves with a keyboard, society loses its mind. The headlines write themselves. Montreal youth accused of using AI to plot school attack. Panic ensues. Pundits clutch their pearls on morning talk shows, demanding immediate bans, stricter guardrails, and algorithmic censorship.
Let me save you the manufactured outrage. Blaming artificial intelligence for a planned violent attack is the exact same lazy panic that blamed heavy metal music in the nineties, video games in the two-thousands, and comic books in the fifties.
It is a comforting delusion. If a chatbot is to blame, adults do not have to examine why a kid felt isolated enough, angry enough, and desperate enough to turn to violence in the first place.
The Algorithmic Boogeyman Fallacy
Let us define what an adolescent text generation model actually is. It is a probabilistic text predictor. It does not possess intent, malice, or a dark master plan. It completes patterns based on a colossal matrix of human writing scraped from the internet.
When a troubled kid inputs a prompt about attacking a building, the system does not magically corrupt an innocent mind. It responds based on the statistical likelihood of what words usually follow those inputs in its training data.
I have spent two decades building and auditing software systems. I have watched executives panic over every new technology wave, throwing millions at compliance theater while ignoring fundamental human failures.
The media wants you to believe that a piece of software acted as a co-conspirator. That is nonsense. The software acted as a mirror. If you do not like what you see in the mirror, breaking the glass changes nothing about your face.
The Real Failure Is Human Abandonment
Every single time a tragedy or near-tragedy occurs involving a minor and digital tools, investigators eventually uncover the same boring, agonizingly human root causes. Total social isolation. Zero parental oversight. Severe mental distress ignored by overloaded school systems. A complete lack of real-world community.
A chatbot does not invite a kid to sit at its lunch table. A machine learning model does not notice when a teenager stops making eye contact, loses weight, or displays sudden behavioral shifts.
Focusing on the prompt history misses the entire point. By hyper-focusing on the technology used to draft a plan, we give a free pass to the parents, teachers, and peers who failed to notice that a human being was slipping into a psychological abyss.
The Dangerous Solution Of Over-Regulation
The knee-jerk reaction from lawmakers to these incidents is predictable. They want age verification locks on everything. They want developers to build proactive snitchware into every consumer application, monitoring every keystroke for signs of deviance.
Imagine a scenario where every single text prompt typed by a minor is flagged, stored, and reviewed by automated safety filters or human moderators. Privacy vanishes completely. A generation grows up under total digital surveillance, convinced that privacy is a myth and expression is a hazard.
Worse, it does not stop violence. Kids who want to bypass restrictions will simply use open-source, locally run models on cheap hardware that have no corporate guardrails whatsoever. You cannot regulate malicious intent out of existence by locking down commercial chat interfaces.
What We Should Be Doing Instead
If we actually care about stopping adolescent violence, we need to stop looking for tech-sector scapegoats.
First, we need to fund mental health resources that actually exist in the physical world, not just through a screen. Second, parents need to reclaim their jobs as supervisors and active participants in their children's inner lives. Handing a smartphone to an unsupervised child and treating it as a digital babysitter is parental negligence, pure and simple.
The machine did not plot an attack. A human mind did. Fix the human.
