Sam Altman, chief of OpenAI, issued a public apology after the company did not alert law enforcement about an account later linked to a mass shooting in Tumbler Ridge that left eight people dead. In a letter shared publicly, Altman acknowledged that the company should have taken stronger action after identifying concerning activity associated with the suspect months before the attack.
He stated that the account had been flagged and banned in June for violating platform policies related to potentially harmful behaviour, but it was not referred to authorities because it did not meet the internal threshold for escalation at the time. In hindsight, Altman admitted that this decision was a failure and expressed regret for not notifying law enforcement earlier.
According to investigators, the February 10 incident involved 18-year-old Jesse Van Rootselaar, who allegedly killed family members at home before carrying out a shooting at a local secondary school, resulting in multiple deaths and injuries. The suspect later died by suicide.
The company’s handling of the account has drawn criticism, particularly over whether earlier intervention could have helped prevent the tragedy. David Eby indicated that there may have been an opportunity for action before the violence occurred, raising broader concerns about the responsibilities of technology platforms in such situations.
Altman said he had spoken with local leaders, including Tumbler Ridge Mayor Darryl Krakowka and Premier Eby, who conveyed the scale of grief and anger within the community. He acknowledged that while an apology cannot undo the harm, it was important to recognise the loss and take responsibility for the lapse.
He also committed to improving coordination with governments and law enforcement agencies to strengthen safeguards and ensure that similar warning signs are handled more effectively in the future. However, Eby described the apology as necessary but insufficient, reflecting ongoing debate over accountability and the role of AI platforms in identifying and reporting potential real-world threats.
