Power & Politics OpenAI CEO agrees to safety changes after Tumbler Ridge: minister
Mar 4, 2026
Michael Chong, Conservative foreign affairs critic, offers a sharp critique of Canada’s stance on U.S. and Israeli strikes. Evan Solomon, Canada’s Minister of AI and Digital Innovation, recounts meetings with OpenAI leadership and safety commitments after the Tumbler Ridge tragedy. They discuss reporting threats to the RCMP, reviews of past flags, and Canada’s diplomatic balancing on Iran.
AI Snips
Chapters
Transcript
Episode notes
Canada Secures Direct RCMP Contact From OpenAI
- Canada pressed OpenAI to change safety processes after the Tumbler Ridge shooter used ChatGPT and the account was shut down without police being notified.
- Evan Solomon secured Sam Altman's agreement to direct RCMP contact and a review of past flagged cases under new criteria.
Verify Safety Claims With Independent Expert Review
- Demand independent verification of company safety claims rather than taking press releases at face value.
- Evan Solomon asked the Canadian AI Safety Institute to do a detailed technical assessment of OpenAI's new threat-detection protocols.
Jurisdictional Blind Spots Led To Missed Escalation
- OpenAI reportedly flags threats to the FBI but previously did not escalate to Canadian law enforcement, creating jurisdictional blind spots.
- Solomon insisted Canadians should assess Canadian threats and secured agreement to include Canadian experts in safety operations.