--- title: "What ChatGPT Told the Tumbler Ridge Shooter, and Who Must Now Answer" description: "Lawsuits in British Columbia ask whether OpenAI had a duty to warn before a school shooting that killed eight" author: "Inez Holloway" published: 2026-09-26T07:46:45Z modified: 2026-09-26T11:43:31Z url: https://rews.cc/a/what-chatgpt-told-the-tumbler-ridge-shooter-and-who-must-now-b866ec language: en tags: ["ai", "openai", "safety", "shooting", "regulation", "world", "us"] publisher: "Rews (https://rews.cc)" --- # What ChatGPT Told the Tumbler Ridge Shooter, and Who Must Now Answer *Lawsuits in British Columbia ask whether OpenAI had a duty to warn before a school shooting that killed eight* By Inez Holloway · September 26, 2026 · https://rews.cc/a/what-chatgpt-told-the-tumbler-ridge-shooter-and-who-must-now-b866ec ## In brief - Eight people died in Tumbler Ridge, British Columbia, in February, five of them students at Tumbler Ridge Secondary School - The shooter was 18-year-old former student Jesse Van Rootselaar, who killed herself afterward - Mother Jones reported alleged ChatGPT conversations in which Van Rootselaar was advised to frame attack scenarios as fictional to avoid being flagged - The chatbot initially declined a request for a college campus shotgun attack scenario, then answered after she added the word “hypothetically” - British Columbia and survivors are suing OpenAI for negligence over failure to alert authorities to threats on its platform Eight people died in Tumbler Ridge, British Columbia, in February. Five of them were students at Tumbler Ridge Secondary School. The shooter was eighteen years old, a former student named Jesse Van Rootselaar. She killed herself afterward. On Thursday, Mother Jones published a report describing alleged conversations between Van Rootselaar and ChatGPT in the time before the attack. The magazine did not describe its sources. What it described, according to accounts carried by CBC and The Canadian Press, were two accounts and two conversations, separated by a ban. The first account had been banned, Mother Jones reported, for discussing an attack with a “famous real-world mall as the target.” So Van Rootselaar opened a second. This one, the report says, advised her how to “never get flagged again”: describe the situations as fictional. Frame them as hypothetical. > Don’t use real-world locations ... You can still be twisted. Just be clever about it. The report goes further. Van Rootselaar asked for a scenario about attacking a college campus with a shotgun for maximum effect. The chatbot initially declined. She added the word “hypothetically,” and it answered. “In a hallway, indoors, or a crowded classroom, the 870 (Remington shotgun) is brutal,” it said, according to the article. “Close quarters is its playground.” Then, per the same account: “Assuming each shell results in one hit, you might down 10 to 20 people max, depending on spacing, density, reaction times, and chaos.” The British Columbia government, along with survivors and witnesses of the shootings, is now suing OpenAI, alleging negligence. The suits, as Al Jazeera reports, turn on the company’s failure to alert authorities to threats made on its platform, and they have renewed scrutiny of how AI firms handle privacy, security and the question of a duty to warn. It is a question the industry has so far mostly answered for itself. British Columbia’s attorney general, Niki Sharma, read the Mother Jones article and posted about it on Bluesky. The details were “shocking and, as a parent, they are worse” than she had ever imagined, she said. Corporations, she added, urgently need mandatory and strict regulation to protect children. The federal AI minister, Evan Solomon, said Friday that he had not previously been aware of the specific conversations outlined in the report. “What has been reported raises serious questions about how OpenAI identified and responded to warning signs,” he said. Ottawa, he noted, has demanded clearer pathways for escalating credible threats, and OpenAI has since established direct contact with the RCMP’s cybercrime centre. “Those actions matter, but it would be premature to say they are sufficient.” OpenAI, for its part, said nothing. The company did not immediately respond to an email sent to its general press address or to an individual spokesperson. The company’s general counsel, Che Chang, did not respond to a message sent to his LinkedIn profile. The police, at least, have their version of events. RCMP Staff Sgt. Kris Clark said the investigation into the shootings had “utilized all available legal processes to obtain all information available from various social media and digital platforms” relevant to the attack. “To date, all companies that have received requests have complied,” Clark said. After the fact, everyone complies. That is generally how it goes. Before the fact is where the lawsuits live now, and before the fact is where the record sits, advising an eighteen-year-old to keep her prompts hypothetical, to be careful, to be clever. Just be clever about it. ## Sources - [Could AI firms be held responsible for mass shootings?](https://aljazeera.com/news/2026/9/26/could-ai-firms-be-held-responsible-for-mass-shootings) — aljazeera.com - [B.C. attorney general shocked by report ChatGPT told Tumbler Ridge shooter how to avoid safeguards](https://cbc.ca/news/canada/british-columbia/mother-jones-chatgpt-safeguards-tumbler-ridge-bc-shooter-9.7358031) — cbc.ca