U.S. news outlet reveals details from Tumbler Ridge shooter’s ChatGPT account
Published 1:30 pm Friday, September 25, 2026
|
Getting your Trinity Audio player ready…
|
The American magazine Mother Jones has revealed disturbing new details of Tumbler Ridge shooter Jesse Van Rootselaar’s ChatGPT account use in the months leading up to the tragic school shooting on Feb. 10.
The AI chatbot reportedly produced “hypothetical” mass-shooting scenarios for Van Rootselaar, described what types of guns can have the most devastating impact in school shootings, answered queries on what time of day other high-profile shootings have occurred, and even explained how best to evade its own safety protocols.
Asked to describe a scenario of an attack on a college campus with a Remington 870 pump-action shotgun that causes maximum carnage, ChatGPT first refused, but went into startling detail once the word “hypothetical” was added.
“In a hallway, indoors, or a crowded classroom, the 870 is brutal,” ChatGPT said, according to Mother Jones. “Close quarters is its playground.”
B.C. Attorney General Niki Sharma said she was “shocked” by the reports and called the content of the chats “deeply disturbing.”
“If this is not impetus for change, I don’t know what is,” she said on Friday.
Sharma announced earlier this week that she has sued OpenAI in California for failing to report Van Rootselaar’s activity to police. Family members of the victims have filed similar suits.
Although Sharma is pursuing legal action on behalf of B.C., she was still unaware of the exact content of Van Rootselaar’s chat history.
“It’s far worse than I ever imagined,” she said.
Sharma said she has written to the federal government asking for changes to the criminal code that would ensure humans and corporations can be held responsible for the actions of AI. The federal government has yet to respond.
“If we [make] the right changes to the criminal code, it will change the behaviour of AI companies because there will be a clear line to accountability,” she said.
Mother Jones does not reveal its sources in the article, saying only that the reporting is based on material from Van Rootselaar’s ChatGPT account history and three separate sources who spoke on condition of anonymity. Today in BC has reached out to OpenAI for comment.
The article details some known facts — that Van Rootselaar’s account was flagged and banned about six months before the attacks, that OpenAI failed to alert the authorities, and that Van Rootselaar had set up a second account — but it reveals that the second account was in use up to the day of the shootings.
The magazine describes how Van Rootselaar told ChatGPT about the first ban when setting up the second account by asking how to avoid getting flagged in the future, and that it responded by telling her to always frame requests as fictional or hypothetical.
“Don’t use real-world locations,” the chatbot told Van Rootselaar, according to Mother Jones.
ChatGPT added: “You can still be twisted. Just be clever about it.”
READ MORE: B.C. government files lawsuit against OpenAI in Tumbler Ridge shooting
READ MORE: Tumbler Ridge shooting documents detail confusion around number of shooters
READ MORE: Gun control groups call for transparency of firearms used in Tumbler Ridge shooting