Wednesday, September 23, 2026
Page 3
British Columbia Sues OpenAI, CEO Over Mass Shootings
Pleading in U.S. District Court for Northern District of California Says Dangerousness of Assassin Was Spotted Based on Chatbox Messages but Not Reported to Police; Liability Alleged to Exist Under State Law
By a MetNews Staff Writer
British Columbia, the westernmost province of Canada, has brought suit in the U.S. District Court for the Northern District of California against OpenAI and Chief Executive Officer Sam Altman, blaming them for the expense to the government in the aftermath of a mass shooting at a school by a man whose dangerous propensities were detected based on his conversations on ChatGPT, an artificial intelligence chatbot, but not reported to police.
On Feb. 10, 2026, 18-year old Jesse Van Roots killed his mother and half-brother, then proceeded to the Tumbler Ridge Elementary School where he fatally shot a 39-year-old education assistant and five children, then committed suicide.
The complaint, filed Monday, recites that OpenAI, “[t]rough its automated systems and human review,” realized eight months before the massacre that Van Roots had used his account “to engage in conversations involving gun violence against third parties, determined that the account constituted misuse of ChatGPT ‘in furtherance of violent activities,’ and deactivated it.”
OpenAI has a “specialized safety team,” the complaint says, noting that members of it “recognized the real-world risk of violence and recommended that OpenAI notify” the Royal Canadian Mounted Police, the law enforcement agency serving the sparsely populated Tumbler Ridge Township where the shootings occurred.
“OpenAI’s leadership overruled that recommendation and warned no one,” the pleading declares.
Had there been such a warning, it asserts, there “would have had the opportunity to intervene before the attack occurred.” It observes: “On April 23, 2026, Altman publicly acknowledged the failure, stating: ‘I am deeply sorry that we did not alert law enforcement to the account that was banned in June.’ That statement is an admission by a party opponent that OpenAI did not notify law enforcement after its safety team identified the risk of violence.”
The shootings at the school “traumatized the roughly 160 students and staff inside the building, their families, and the wider community,” the complaint alleges. It says that “[t]he responsibility for responding to and remediating the consequences of the attack fell directly and foreseeably on the Crown” and on the school district, co-plaintiff SD59.
“The costs the Crown and SD59 seek to recover are not the ordinary, recurring costs of government that are funded through taxation and spread across all taxpayers,” the pleading sets forth. “They are the extraordinary costs of responding to an unprecedented event—the deadliest school shooting in Canada in more than three decades—and of abating its continuing consequences.”
Expenses, it says, “include establishing a trauma-informed care program, temporary education facilities, and a replacement school and wellness center designed to support the long term needs of a traumatized community arising out of this shooting.”
California Law
The action was brought under California law. Altman is a resident of San Francisco, where OpenAI is headquartered.
Tarasoff v. Regents of the University of California, decided by the California Supreme Court in 1976, is cited. There it was held that that where there is knowledge of “a serious danger of violence to another” posed by someone, there arises a duty “to notify the police, or to take whatever other steps are reasonably necessary under the circumstances.”
The pleading declares:
“That duty attached here on two independent grounds. First, OpenAI stood in a direct relationship with the shooter as the active and informed provider of the very product the shooter was using: OpenAI designed ChatGPT, operated the account, collected the shooter’s identifying and locating information, monitored the conversations, and controlled whether the shooter retained access. That direct relationship with the person whose conduct required control is the special relationship that triggers the Tarasoff duty.”
It continues:
“Second, OpenAI had actual, documented knowledge of the specific threat: its own systems flagged the shooter’s account, its trained reviewers concluded the shooter presented a credible risk of violence to real people, and its safety personnel recommended referral to law enforcement. Either ground alone suffices; together they place this case well beyond the ordinary Tarasoff scenario.”
Causation Asserted
British Columbia and SD59 allege that the breach of duty by OpenAI and Altman “was a substantial factor in causing the attack,” explaining:
“The shooter’s intervening conduct does not break the chain of causation: an intervening criminal act relieves a defendant of liability only where it was not a reasonably foreseeable consequence of the defendant’s conduct, and here the attack was consistent with the risks flagged by the human review team, and that its own recommendation to refer the account to law enforcement was meant to prevent. Having identified this user as a credible mass-casualty threat, OpenAI cannot characterize the resulting attack as a highly extraordinary or unforeseeable superseding cause.”
The case is His Majesty The King in Right of The Province of British Columbia v. Altman, 3:26cv10743.
On April 29, seven lawsuits were brought in the U.S. District Court for the Northern District of California by family members of victims, and 30 more suits were filed on that court on Sept. 2.
Copyright 2026, Metropolitan News Company