The Globe and Mail reports in its Monday, Feb. 23, edition that two artificial intelligence policy experts urge that the upcoming federal on-line harms bill must include regulations for AI chatbots and establish a framework for reporting credible threats, following revelations that concerning content from the Tumbler Ridge shooter was flagged but not reported to authorities. The Globe's Marie Woolf writes that earlier this month, the 18-year-old shooter killed five children and a teacher's aide at her former B.C. school after killing her mother and half-brother at her home. Her posts were flagged by Microsoft-backed OpenAI screening systems, leading to a suspension of her ChatGPT account due to concerning content. However, OpenAI did not inform law enforcement in June, citing no "credible or imminent planning." The experts recommend that the on-line harms bill include guidelines for AI platforms on reporting violent content to police and emphasize the need for transparency in their risk mitigation policies, especially regarding children's safety. McGill University's Helen Hayes says, "Internal flagging without a clear, legally defined escalation path is insufficient."