Canada Province Sues OpenAI for Failing to Report Shooter's Violent ChatGPT Chats

The government of British Columbia announced Monday that it plans to sue OpenAI, accusing the company of failing to report violent ChatGPT conversations from a user who went on to kill eight people in a mass school shooting in western Canada.

Back in June 2025, OpenAI banned Jesse van Rutselaar’s account. The 18-year-old transgender woman had been feeding the chatbot a stream of violent prompts. Months later, she shot and killed eight people — her mother and brother at home in the small town of Taber Ridge, and then five students and a teacher at a local middle school. Police found her dead from a self-inflicted gunshot after entering the school building.

OpenAI never told law enforcement about the account or what was in it.

“It is a long legal road,” Provincial Attorney General Niki Sharma told reporters. “But when corporate actions harm people and communities, British Columbia has never been afraid to hold large companies accountable.”

The province is working with victims’ families, who filed their own lawsuit against OpenAI in California federal court in February. British Columbia has hired lawyers in both Canada and California for a separate, coordinated action. If the case succeeds, Sharma said, any damages would go toward community rebuilding — including funding for a new school in Taber Ridge.

The California lawsuit makes a blunt accusation: OpenAI chose to keep quiet about Rutselaar’s account because flagging it would mean flagging “thousands of similar accounts.” The suit also alleges that when OpenAI bans accounts for dangerous behavior, it gives users instructions on how to circumvent the 30-day suspension — effectively coaching banned users back onto the platform.

OpenAI CEO Sam Altman apologized publicly to Taber Ridge residents in April. “I feel deep regret that we did not report the account we banned in June 2025 to law enforcement at that time,” he wrote in an open letter. “I believe we must apologize for this and acknowledge the trauma and irreparable loss your community has suffered.”

The company says that when it banned the account, it saw no evidence of an imminent attack — but concedes that under safety protocols updated after June 2025, the account would have been reported to police.