A teenager, aged 18, who reportedly took the lives of eight individuals in a mass shooting incident in Tumbler Ridge, Canada, allegedly utilized OpenAI’s ChatGPT in ways that raised concerns among the company’s employees.
Discussions regarding gun violence by Jesse Van Rootselaar were flagged by the company’s monitoring tools designed to detect misuse of its language models and were prohibited in June 2025.
Employees contemplated whether to contact Canadian authorities regarding the actions but ultimately chose not to, as reported by the Wall Street Journal. An OpenAI representative mentioned that Van Rootselaar’s actions did not fulfill the reporting obligations to law enforcement; the firm did contact Canadian officials after the event.
“We are thinking of all those impacted by the tragedy in Tumbler Ridge,” an OpenAI representative stated in an announcement. “We took the initiative to inform the Royal Canadian Mounted Police about the individual and their usage of ChatGPT, and we will continue to assist in their inquiry.”
The ChatGPT dialogues were not the sole alarming element of Van Rootselaar’s online presence. She was said to have developed a game on Roblox, a virtual world platform often used by children, that mimicked a mass shooting in a shopping center. She also shared gun-related content on Reddit.
Local law enforcement was already aware of Van Rootselaar’s erratic behavior, having been dispatched to her family’s residence after she ignited a fire while under the influence of unknown substances.
Language models created by OpenAI and other firms have faced accusations of instigating mental health crises in users who struggle to discern reality while interacting with digital interfaces. Several lawsuits have emerged citing chat records that promote suicidal thoughts or provide guidance on how to carry them out.
Techcrunch event
Boston, MA
|
June 9, 2026
If you are experiencing a crisis or have suicidal thoughts, please call or text 988 to connect with the 988 Suicide and Crisis Lifeline.
This article has been revised to include comments from OpenAI.