OpenAI detected violent intent in ChatGPT user before Canadian mass shooting, did not alert authorities
OpenAI reportedly detected that a ChatGPT user was describing violent scenarios in June 2025 and considered alerting Canadian police but concluded there was no credible imminent risk and did not report the account. Several months later, the user allegedly carried out a mass shooting in British Columbia that killed nine people.
What happened
In June 2025, OpenAI's internal abuse detection systems flagged the ChatGPT account of Jesse Van Rootselaar as supporting or facilitating violent activities after the user described scenarios involving firearm violence over several days.
OpenAI representatives evaluated whether to report the account to Canadian police but decided that no credible and imminent risk of serious violence existed, which was the threshold required to alert authorities. The account was closed in June 2025 for violating usage policies.
Reported impact
A week after the article date of 21 February 2026, Jesse Van Rootselaar killed nine people in Tumbler Ridge, British Columbia, including a teacher and multiple children at a school and family members at home.
Governance controls
After the incident, OpenAI reported that it contacted police and provided investigators with information about the suspect's use of the platform. The company stated it would continue to cooperate with authorities during the investigation.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.