OpenAI flagged violent chats, didn't alert police
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Jesse Van Rootselaar discussed gun‑violence scenarios with ChatGPT in June, triggering the model’s automated review system.
- OpenAI employees raised concerns that the posts could foreshadow real‑world violence and urged leaders to contact authorities.
- OpenAI decided the content did not constitute an “imminent and credible risk” and chose not to refer the account to law enforcement.
- OpenAI subsequently banned Jesse Van Rootselaar’s ChatGPT account but took no further precautionary actions at the time.
- Royal Canadian Mounted Police received proactive outreach from OpenAI with information on the individual and their use of ChatGPT.
- Tumbler Ridge Secondary School was the site of the February 10 shooting that left nine dead, 27 injured, and the suspect dead from a self‑inflicted wound.
- Kayla Wood expressed sympathy for victims and said OpenAI will continue to support the RCMP investigation.
Why it matters: Victims’ families and the broader public miss a chance for early intervention, while OpenAI faces scrutiny for balancing privacy with safety, highlighting the tension between AI‑driven content monitoring and law‑enforcement cooperation. The decision not to notify police before the February 10 shooting underscores the trade‑off between user confidentiality and public safety, affecting public trust in AI platforms.
