A mass shooting in British Columbia has drawn attention to OpenAI CEO Sam Altman's push for privacy protections for AI conversations. The shooter reportedly discussed gun violence scenarios with ChatGPT months before the attack, but OpenAI did not alert authorities. Canadian officials are questioning the company's handling of the matter.
On February 10, Jesse Van Rootselaar, described as wearing a dress, killed his mother and brother before proceeding to Tumbler Ridge Secondary School in British Columbia, where he fatally shot six more people, including five children aged 12 and 13. Van Rootselaar died that day from a self-inflicted gunshot wound.
Months prior, Van Rootselaar had conversations with ChatGPT about gun violence scenarios in June 2025, which raised concerns among OpenAI employees. According to a Wall Street Journal report, these interactions were not reported to law enforcement, though his account was banned. The content of the conversations remains unclear.
After identifying Van Rootselaar as the perpetrator, OpenAI contacted the Royal Canadian Mounted Police to assist the investigation. However, Canadian officials expressed dissatisfaction with OpenAI's response and have summoned company employees for discussions on the incident.
The event has spotlighted comments by OpenAI CEO Sam Altman from a September interview with Tucker Carlson, where he advocated for an 'AI privilege' policy. Altman stated, “If I could get one piece of policy passed right now, relative to AI, the thing I would most like... is I’d like there to be a concept of AI privilege.” He compared it to protections for doctor-patient or lawyer-client communications, arguing that society has an interest in keeping such AI interactions private from government access, even via subpoena.
Altman noted he had recently advocated for this in Washington, D.C., expressing optimism about adoption.
British Columbia Premier David Eby commented on reports of OpenAI receiving potential advance notice of the shooter's intentions. “With shock and dismay, like many British Columbians, I am trying to figure out how it could be possible that a large group of staff within an organization could bring this kind of information forward and ask the police to be called and the decision be made not to do that,” Eby said. From outside, he suggested OpenAI might have prevented the shooting and urged Canada's federal government to set a national reporting threshold for AI firms on violence plots.
Canada’s Federal AI Minister Evan Solomon met with OpenAI staff on Tuesday to review safety protocols. OpenAI maintains its models discourage real-world violence and include systems to flag troubling content for review and possible law enforcement referral. The company did not comment on whether Altman continues to support AI privacy immunity.
Such privileges could complicate responses to violence threats, similar to mandatory reporting duties for mental health professionals in cases of imminent danger.
