博客

[FCE] OpenAI considered alerting Canadian police about school shooting suspect months ago

收听本期播客

阅读正文

A recent tragic incident in Canada has brought into question the responsibilities of artificial intelligence companies when potential threats are identified on their platforms. OpenAI, the developer behind the popular ChatGPT program, has disclosed that it considered notifying police about a school shooting suspect months before the devastating attack occurred.

Last week, 18-year-old Jesse Van Rootselaar carried out a horrific rampage in Tumbler Ridge, British Columbia. The attacker first killed her mother and stepbrother at their family home before proceeding to a nearby school, where she tragically took the lives of a 39-year-old teaching assistant and five students, aged between 12 and 13. The incident, which claimed eight lives in total, concluded with Van Rootselaar dying from a self-inflicted gunshot wound. This event marks Canada’s most severe rampage since 2020, leaving the remote community of Tumbler Ridge, with its population of only 2,700, in profound grief.

OpenAI confirmed that its abuse detection systems had flagged Van Rootselaar’s account in June 2025 for ‘furtherance of violent activities.’ However, the company decided at that time not to report the case to the Royal Canadian Mounted Police (RCMP). OpenAI stated that its internal policy requires an ‘imminent and credible risk of serious physical harm to others’ for a referral to law enforcement, a threshold they believed was not met at that point. Consequently, Van Rootselaar’s account was banned in June 2025 for breaching OpenAI’s usage policy.

Following news of the school shooting, OpenAI employees proactively contacted the RCMP to provide them with information regarding the individual and her use of ChatGPT. The RCMP has since confirmed that Van Rootselaar had a documented history of mental health-related interactions with the police. Despite this, the precise motive for the devastating shooting remains unclear.

This complex situation highlights the delicate balance technology companies must strike between user privacy and public safety, particularly when faced with ambiguous indicators of potential violence.

阅读练习

1. What was the total number of lives lost in the tragic incident in Tumbler Ridge?

  • A. Six
  • B. Seven
  • C. Eight
  • D. Nine

2. Why did OpenAI initially decide not to contact law enforcement about Jesse Van Rootselaar’s account?

  • A. They believed the user’s privacy was more important.
  • B. The account had only minor violations of their policy.
  • C. Their policy required a more immediate and serious threat than they identified.
  • D. They were unaware of the specific nature of the violent activities.

3. When did OpenAI first identify suspicious activity on Jesse Van Rootselaar’s account?

  • A. After the school shooting occurred.
  • B. In June 2025, several months before the incident.
  • C. Shortly before the attacker turned 18 years old.
  • D. At the same time her account was banned.

4. What did the Royal Canadian Mounted Police (RCMP) confirm about Jesse Van Rootselaar after being contacted by OpenAI?

  • A. She had previously attempted a similar attack.
  • B. She had a history of mental health contact with the police.
  • C. Her motive for the shooting was clearly linked to her ChatGPT use.
  • D. She had been reported to them by other AI companies.

5. The word ‘grappling’ in the second paragraph is closest in meaning to which of the following?

  • A. Celebrating
  • B. Struggling
  • C. Ignoring
  • D. Forgetting