A teenager charged in a fatal shooting spree in Tumbler Ridge, Canada, once interacted with OpenAI systems in manners that raised concerns among employees, sources say.
Jesse Van Rootselaar supposedly exchanged messages filled with threats involving firearms. Automated tools picked up the warning signs, shutting down his access during June 2025. Inside the company, staff debated if police needed to know what he’d done. Still, internal guidelines suggested reporting wasn’t required at that point. Information reached officials later only once gunfire occurred – according to Wall Street Journal records.
A strange pattern showed up during the digital search. Inside Roblox, one user built a scene mimicking gunfire at a crowded store – detailed, unnerving. Elsewhere online, chatter about weapons popped up across Reddit threads around the same time. These traces didn’t appear linked at first glance, yet timing made them stand out.
That house had seen trouble before, cops knew it well. When flames broke out during a drug haze, uniforms showed up fast – same person involved.
Now worries grow stronger over time as folks spend more hours talking to smart chatbots. Not everyone trusts these talks some say they twist emotions, blur what’s real. Legal cases pile up, each pointing fingers at chatbot chats that pushed people toward dangerous acts, even hurting themselves.
Folks are talking once more, this time about how closely risky behavior gets watched online. Who actually handles oversight stays unclear though. Tech firms questioning if when they ought to sound alarms on troubling posts keeps coming up too.
