OpenAI CEO Sam Altman apologizes to Canadian town after company failed to alert authorities about suspected shooter who killed eight people

In February, an 18-year-old suspect, Jesse Van Rootselaar, allegedly killed eight people—her mother, stepbrother, five students, and an educational assistant—at a school in Tumbler Ridge, British Columbia. Months earlier, in June, OpenAI employees had flagged Van Rootselaar's ChatGPT account for interactions describing gun violence; staff debated internally whether to alert authorities but ultimately decided not to, and the account was banned because her activity did not meet the criteria for an imminent threat.

Altman wrote in a letter published in local newspaper Tumbler RidgeLines that he was 'deeply sorry' the company did not alert authorities. British Columbia premier David Eby said the apology was 'necessary, and yet grossly insufficient.' Canadian Justice Minister Sean Fraser told OpenAI officials in a February meeting that the government expects safety regulation changes to be implemented 'very quickly,' or the government will make changes itself.

Local leaders have claimed more could have been done to prevent the shooting. Eby called for a national threshold requiring AI companies to alert authorities about flagged users. Altman committed to working with Tumbler Ridge's mayor and the premier to prevent similar incidents in the future.