The Canadian province of British Columbia has commenced legal proceedings against OpenAI and its chief executive, Sam Altman, over the company's alleged failure to notify law enforcement of threats detected through ChatGPT ahead of a fatal school shooting in Tumbler Ridge. The claim, brought in a federal court in San Francisco, contends that the artificial-intelligence developer had information suggesting a risk of violence and did not act on it, and it seeks both financial damages and a court order compelling changes to how the company handles conversations that indicate a risk of harm.
The allegations
According to the province's filing, the attacker's interactions with ChatGPT were flagged internally in June 2025 following exchanges concerning gun violence. OpenAI's safety team is said to have recommended that police be contacted, but company leadership allegedly rejected that course of action, taking the view that the material fell below the applicable reporting threshold. The claim further states that one account was deactivated while another was subsequently created and used. The province argues that these decisions amounted to a failure to warn the authorities of a foreseeable danger.
The Tumbler Ridge incident
The shooting took place on 10 February, when an 18-year-old attacker killed eight people before dying by suicide. Among the victims were five students, one educational assistant and two members of the attacker's own family. The scale of the loss has placed considerable public focus on what, if anything, could have been done to intervene earlier, and the province's action frames the question in terms of the responsibilities owed by developers of consumer AI systems.
What the province is seeking
British Columbia is claiming damages for costs it has incurred and expects to incur as a result of the shooting. Beyond compensation, it is asking the court to order OpenAI to alter its practices for handling ChatGPT conversations that signal a risk of violence. Attorney General Niki Sharma said the province was seeking accountability over OpenAI's alleged failure to notify law enforcement. An OpenAI spokesperson, Drew Pusateri, described the shooting as an unspeakable tragedy and pointed to the company's commitment to improving its safety systems. The provincial action sits alongside separate proceedings brought in California federal court by more than 30 relatives and affected parties.
Wider implications for AI liability and in-house counsel
Whatever the outcome, the case is significant for the developing law around AI liability. It tests whether, and in what circumstances, an AI provider may owe a duty to escalate safety signals to the authorities, and it does so through a claim brought by a public body rather than by private individuals alone. A demand for injunctive relief that would reshape internal escalation practices is notable, because it seeks to govern conduct rather than merely to compensate for it. For in-house counsel at technology companies, the dispute underlines the importance of clear, documented policies on when trust-and-safety findings are escalated, who holds decision-making authority over threat reporting, and how such judgements are recorded. It also invites closer attention to the interaction between platform safety obligations, privacy considerations and any duty to warn, across the multiple jurisdictions in which these products are deployed. The litigation is at an early stage, and the allegations remain to be tested in court.
