British Columbia Sues OpenAI Over School Shooting

Published: 22 September 2026. The English Chronicle Desk. The English Chronicle Online.

The Canadian province of British Columbia has filed a lawsuit against OpenAI and its chief executive, Sam Altman, alleging that the deadly school shooting in Tumbler Ridge could have been prevented if the company had alerted law enforcement about the shooter’s use of ChatGPT to discuss and plan violence.

The lawsuit was filed in a federal court in San Francisco on Monday and names OpenAI and Altman as defendants. British Columbia is seeking financial damages to help cover the extensive costs arising from the February attack, including recovery efforts, mental health services and the replacement of the school where the shooting occurred. The province is also asking the court to require changes to the way OpenAI handles ChatGPT conversations that may indicate a risk of serious violence.

Nine people, most of them children, were killed in the Tumbler Ridge attack. According to police accounts cited in the lawsuit, the shooter, Jesse Van Rootselaar, first shot her mother and stepbrother at their home before going to her former school on 10 February. There, she killed an educational assistant and five students aged 12 and 13. Van Rootselaar, who was 18 at the time, later died by suicide.

The lawsuit centres on what British Columbia says OpenAI knew about Van Rootselaar’s online interactions before the attack and how the company responded to those warning signs. The province alleges that OpenAI’s safety team had identified concerning conversations involving gun violence months before the shooting and that members of the safety team recommended notifying police.

According to the allegations contained in the lawsuit, those recommendations were not acted upon after being overruled by senior OpenAI leadership, including Altman. The claims are based in part on accounts from OpenAI whistleblowers who spoke publicly after the shooting. These allegations have not been established as facts by a court, and OpenAI has disputed the broader claims made against it.

Van Rootselaar’s interactions with ChatGPT were reportedly first flagged by OpenAI in June 2025. The company subsequently deactivated the account, but the lawsuit alleges that she was able to create another account and continue using the platform. British Columbia argues that the continued access was significant because the province believes the company had already identified warning signs associated with violent activity.

OpenAI has offered a different account of its internal assessment. After reports about the case emerged, the company said the account had been identified by systems designed to detect misuse of its models in connection with violent activities. However, OpenAI said the circumstances did not meet its internal threshold for reporting the conversations to law enforcement.

The dispute therefore raises a difficult question about when technology companies should involve police when users appear to be discussing potential violence. OpenAI has said that its systems are designed to refuse requests that could meaningfully enable violence and that law enforcement is notified when conversations indicate an imminent and credible risk of harm to other people.

British Columbia’s lawsuit seeks to challenge aspects of that approach. The province argues that the company should have taken stronger action in the Tumbler Ridge case and is asking for changes to the handling of conversations that could lead to violent attacks.

The consequences of the shooting have extended well beyond the immediate loss of life. The school where the attack occurred is being demolished, while British Columbia has taken on the cost of constructing a replacement facility. The province also says it has incurred substantial expenses related to mental health care and other services required after the tragedy.

For families and the wider Tumbler Ridge community, the legal action comes months after the attack as questions continue over whether earlier intervention could have changed what happened. The lawsuit places the conduct of an artificial intelligence company at the centre of those questions, particularly its responsibility when its systems detect potentially dangerous behaviour.

Altman previously addressed the community and expressed deep regret that OpenAI had not contacted law enforcement before the shooting. He promised reforms following the tragedy. British Columbia’s lawsuit alleges that those promised changes were not subsequently implemented despite attempts by the province’s attorney general to engage with the company.

OpenAI spokesperson Drew Pusateri described the shooting as an unspeakable tragedy and said the company remained committed to working with government and law enforcement officials while continuing its safety work.

The British Columbia case is not the first legal challenge to OpenAI arising from the Tumbler Ridge shooting. More than 30 relatives of victims and other people affected by the attack had already filed lawsuits in California earlier in September. Those cases similarly accuse OpenAI of failing to act on opportunities that, according to the plaintiffs, could potentially have helped prevent the attack.

The legal disputes form part of a broader series of cases examining the responsibilities of artificial intelligence companies when their products are allegedly connected to harmful behaviour. In recent months, OpenAI has faced lawsuits in state and federal courts over claims involving violence and other safety concerns.

Florida also brought a lawsuit against OpenAI in June, becoming the first US state to sue the company over alleged safety risks. The case accused OpenAI of misrepresenting the safety of ChatGPT and alleged that the platform had contributed to harm involving children, including claims related to school shootings and self-harm. OpenAI has denied those allegations.

The cases could have implications beyond the individual circumstances surrounding Tumbler Ridge. Artificial intelligence systems are increasingly used for research, communication and planning, making the question of how platforms respond to potentially dangerous conversations a growing issue for regulators, governments and technology companies.

One central issue is the distinction between troubling or disturbing conversations and a genuine, imminent threat. Technology companies must determine what information should remain private, what activity violates platform rules and when circumstances justify involving law enforcement. Those decisions can involve complex questions about privacy, free expression, user safety and the prevention of serious crime.

The British Columbia lawsuit seeks to make those responsibilities part of a legal examination. Rather than focusing only on the actions of the individual shooter, the province is asking whether the company that operated the technology should also bear responsibility for failing to escalate the warning signs it had detected.

The outcome of the case will depend on evidence presented in court and on how judges interpret the legal responsibilities of artificial intelligence companies. The allegations in the lawsuit remain contested, and filing a case does not establish liability.

For the Tumbler Ridge community, however, the legal proceedings represent another stage in the effort to address the consequences of one of Canada’s deadliest recent school attacks. The deaths of nine people, including several children, continue to shape the community’s recovery, while the destruction and rebuilding of the school underscore the lasting impact of the incident.

The case also highlights how rapidly the development of artificial intelligence is creating new challenges for public safety. Companies developing powerful conversational systems must increasingly confront questions about how their technology can be misused and what responsibilities arise when internal safety systems detect potentially dangerous activity.

As British Columbia’s lawsuit moves through the US federal court system, the allegations against OpenAI and Altman will face judicial scrutiny. The proceedings could ultimately contribute to a wider debate over how artificial intelligence platforms should identify, respond to and report credible threats of violence while balancing those responsibilities against privacy and other legal protections.