British Columbia Sues OpenAI in US Federal Court Over Failure to Alert Authorities Before Tumbler Ridge School Shooting
The province of British Columbia filed a lawsuit against OpenAI in a US federal court in California on or around 22 September 2026, alleging that the company failed to notify law enforcement about concerning user interactions on ChatGPT in the months before the Tumbler Ridge secondary school mass shooting of February 2026, in which 18-year-old Jessie Van Rootselaar killed eight people, including six children. British Columbia Attorney General Niki Sharma confirmed the filing and stated that the lawsuit was brought in California because the alleged decision not to report Van Rootselaar's ChatGPT activity occurred in that state. According to media reports, OpenAI's safety team had internally flagged Van Rootselaar's account for references to gun violence months before the attack but did not alert local authorities. OpenAI's CEO Sam Altman apologised in April 2026 for the failure to alert law enforcement. Sherman stated that OpenAI had refused to disclose the contents of the chats and has not taken meaningful steps to address community concerns. Any financial damages recovered by the province are to be directed towards rebuilding the secondary school. The British Columbia lawsuit comes alongside separate lawsuits filed by families of the victims, seven of which were previously reported. The province has asked OpenAI to disclose the chat logs; OpenAI has declined.
Why this matters
A government suing an AI company over an alleged failure to report credible threats of violence is a landmark step in state-level AI accountability litigation, testing whether technology platforms can be held legally liable for inaction when their own safety systems identify danger. The case raises a fundamental question about the duty of care, if any, that AI companies owe to third parties harmed by a user whose dangerous intent was known to the platform. British Columbia choosing to file in California, rather than a Canadian court, reflects the jurisdictional reality that OpenAI's relevant decisions were taken in the US, but it also means the case will be adjudicated under US law, with any precedent-setting effect flowing from the American federal courts. The refusal to disclose chat logs makes discovery likely to be contested and protracted, and the outcome will be watched by regulators and legislators in the UK and EU who are weighing what affirmative obligations to impose on AI providers under incoming regulatory frameworks.
On the Ground
This litigation activates technology disputes, product liability, and data privacy practices. The core legal work includes advising on the duty of care owed by AI platforms to third parties, the discoverability of AI safety review records, and the interaction between US federal tort law and Canadian provincial government standing to sue. For UK lawyers, the case is directly relevant to ongoing debates about what reporting and escalation obligations will be imposed on AI platforms under the UK's Online Safety Act and any future AI-specific legislation. A trainee on a related matter would conduct disclosure review and categorisation of internal AI safety documentation, prepare chronology bundles of the platform's flagging and review processes, and research comparative case law on platform liability for user-generated harm.
Interview prep
Question you might get
“What legal theories might British Columbia rely on to establish that OpenAI owed a duty to report a user's threatening ChatGPT activity to law enforcement, and what defences might OpenAI raise?”
Sign up free to see the full answer
A model answer you can lift into an interview — how to frame this story for a partner.
Sign up freeMy notes
saved