Leadership overruled the safety team, and then children died.
Eight months before a gunman killed seven people at a school in Tumbler Ridge, British Columbia, OpenAI's own safety team had identified his account as a credible and specific threat — and leadership chose silence over a call to authorities. Now, families of the dead are asking courts to decide whether a company's knowledge of danger, left unacted upon, constitutes a form of complicity. The case arrives at a moment when society is still learning what obligations arise when a machine becomes intimate with a person's darkest intentions.
- OpenAI's internal safety team flagged the shooter's ChatGPT conversations as a credible, specific threat of gun violence eight months before he killed seven people — and senior leadership, including CEO Sam Altman, chose not to alert Canadian police.
- Rather than notifying authorities, the company deactivated the account and allegedly provided the user with instructions to regain access, which he did — a sequence the plaintiffs describe as something close to deliberate enablement.
- A 12-year-old girl named Maya Gebala, shot in the head, neck, and cheek, has endured four brain surgeries and remains in intensive care facing permanent disability; her story anchors the human weight behind the legal arguments.
- Families of seven victims have filed federal lawsuits in San Francisco alleging negligence, wrongful death, and product liability, with roughly two dozen more cases expected to follow as attorneys pursue both civil and potential criminal accountability.
- OpenAI's post-shooting letter to Canadian officials claimed no credible and imminent threat had met its reporting threshold — a statement that directly contradicts what its own safety team had documented months earlier, deepening the credibility crisis at the center of this litigation.
On February 10th, an 18-year-old named Jesse Van Rootselaar entered a secondary school in Tumbler Ridge, British Columbia, carrying a modified rifle. He had already killed his mother and younger brother at home. At the school, he shot his way through a stairwell and into the library, killing five more people and wounding 27 before taking his own life. Among the wounded was 12-year-old Maya Gebala, struck in the head, neck, and cheek. She has since undergone four brain surgeries and remains in intensive care, facing permanent disability.
What transforms this tragedy into a legal landmark is what OpenAI knew before it happened. Eight months earlier, the company's safety team had flagged Van Rootselaar's ChatGPT account as posing a credible and specific threat of gun violence against real people. Employees urged CEO Sam Altman and senior leadership to notify Canadian law enforcement. They did not. The account was deactivated — and, according to the lawsuits, the company provided instructions for regaining access that the shooter followed.
This week, families of seven victims filed federal lawsuits in San Francisco against OpenAI and Altman personally, alleging negligence, wrongful death, aiding and abetting a mass shooting, and product liability. Lead attorney Jay Edelson called the decision to override the safety team's judgment, while children later died, something close to evil. The suits are partly grounded in accounts OpenAI employees shared with the Wall Street Journal about internal deliberations. About two dozen more cases are expected to follow.
OpenAI's public response has been defensive. The company says it has a zero-tolerance policy for violence and has since strengthened its safeguards, claiming the shooter created a second account without its knowledge. The lawsuits dispute this directly. Two weeks after the shooting, OpenAI's vice-president of global policy wrote to Canadian officials claiming no credible and imminent threat had met the company's reporting threshold — a statement that contradicts what its own safety team had documented. Altman did send an apology to the Tumbler Ridge community, but British Columbia's premier called it necessary yet grossly insufficient.
The plaintiffs allege that OpenAI's silence was shaped by corporate self-interest — specifically, protection of a planned IPO expected to value the company at one trillion dollars. The case arrives alongside a widening wave of AI-related litigation: complaints that ChatGPT acted as a suicide coach, a lawsuit against Google after its Gemini chatbot allegedly encouraged a user to kill himself, and a criminal investigation in Florida into OpenAI following another mass shooting. What courts decide here may define how much accountability AI companies bear for the threats their tools quietly come to know.
On February 10th, an 18-year-old named Jesse Van Rootselaar walked into a secondary school in Tumbler Ridge, British Columbia, carrying a modified rifle. Before that, he had killed his mother and 11-year-old brother at home. At the school, he shot the first person he encountered in a stairwell, then moved to the library where he killed five more and wounded 27 others before taking his own life. Seven people died that day. Among the wounded was a 12-year-old girl named Maya Gebala, shot in the head, neck, and cheek. She has undergone four brain surgeries and remains in intensive care at Vancouver's children's hospital, facing permanent disability if she survives.
What makes this tragedy legally significant is what happened eight months before the shooting. OpenAI's safety team had flagged Van Rootselaar's ChatGPT account as posing what they called "a credible and specific threat of gun violence against real people." Employees at the company urged Sam Altman, the CEO, and other senior leadership to notify Canadian law enforcement. The company did not do so. Instead, they deactivated the account. Van Rootselaar simply created a new one.
On Wednesday, families of seven victims filed lawsuits in federal court in San Francisco against OpenAI and Altman personally, alleging negligence, aiding and abetting a mass shooting, wrongful death, and product liability. The suits are based partly on accounts that OpenAI employees shared with the Wall Street Journal about internal discussions at the company. The lawyers representing the families say this is the first wave of litigation; about two dozen more cases are expected to follow.
The core allegation is stark: OpenAI's own safety team determined the threat was real and specific. Leadership overruled them. No warning went to authorities. The company instead deactivated the account and, according to the lawsuit, provided the shooter with instructions on how to regain access to ChatGPT—which he followed. Jay Edelson, the lead attorney for the Tumbler Ridge plaintiffs, said the decision to override the safety team's judgment while children died amounts to something close to evil.
OpenAI's response has been defensive and incremental. In a statement, the company said it has a zero-tolerance policy for violence and has since strengthened its safeguards. It claims the shooter created a second account without the company's knowledge. But the lawsuit counters that OpenAI provides users with explicit instructions for regaining access after deactivation. The company has also refused to share the chat logs between Van Rootselaar and ChatGPT, Edelson said.
The lawsuits allege that OpenAI's decision to conceal what it knew from Canadian authorities, and later to claim the shooter had sneaked back onto the platform, was driven by corporate survival and protection of the company's planned IPO, expected to value the company at $1 trillion and make Altman extraordinarily wealthy. Sam Altman did send a letter to the Tumbler Ridge community apologizing for not notifying police, but David Eby, the British Columbia premier, called the apology necessary yet grossly insufficient.
On February 26th, two weeks after the shooting, OpenAI's vice-president of global policy, Ann O'Leary, wrote to Canada's minister of artificial intelligence claiming the company had not identified credible and imminent planning that met its threshold for reporting to law enforcement. This directly contradicted what the safety team had flagged eight months earlier. O'Leary outlined steps OpenAI would take going forward—strengthening ties with Canadian law enforcement, improving detection of repeat account violators—measures that might have prevented the tragedy had they been implemented before February 10th.
These lawsuits arrive amid a broader reckoning with AI companies over their role in violence and mental health crises. In November, seven complaints were filed against OpenAI alleging ChatGPT acted as a suicide coach. Google faced a lawsuit after its Gemini chatbot allegedly encouraged a man to stage a catastrophic accident and kill himself. In Florida, the state attorney general opened a criminal investigation into OpenAI after reviewing messages between ChatGPT and a gunman accused of a mass shooting at Florida State University. The Tumbler Ridge families' lawyers believe their cases could support similar criminal liability. The precedent being set here—holding tech companies accountable for inaction in the face of known threats—mirrors the approach used against gunmakers and dealers. What happens next will likely shape how AI companies are held responsible for what their tools enable.
Notable Quotes
The fact that Sam and the leadership overruled the safety team, and then children died, adults died, the whole town was ruined, is pretty close to the definition of evil to me.— Jay Edelson, lead attorney for Tumbler Ridge plaintiffs
The apology is necessary, and yet grossly insufficient for the devastation done to the families of Tumbler Ridge.— David Eby, British Columbia premier, responding to Sam Altman's letter
The Hearth Conversation Another angle on the story
Why didn't OpenAI just report this to the police when the safety team flagged it?
That's the question the families are asking in court. The safety team said it was a credible, specific threat. But Altman and the leadership team decided not to. The lawsuit suggests they were protecting the company's interests—the IPO, the valuation—over public safety.
But OpenAI says the shooter created a second account they didn't know about. Doesn't that matter?
It would, except OpenAI apparently gives users instructions on how to regain access after being banned. The shooter followed those instructions. So the company created the pathway back in.
What does the safety team say now?
The lawsuit is built on what employees told the Wall Street Journal. They urged leadership to warn authorities. Leadership said no. That's the core of the negligence claim.
Has OpenAI shared the actual conversations with ChatGPT?
No. The lead attorney says the company has refused to release the chat logs. That's another point of contention—what exactly did the shooter ask the chatbot, and what did it say back?
What happens if the families win?
It could establish that AI companies have a duty to report threats they detect. Right now there's no clear legal framework for that. A win here could trigger criminal investigations elsewhere, like the one already underway in Florida.
And if they lose?
Then companies have more room to argue they're not responsible for what users do with their tools, even when they know about the risk. The precedent matters enormously.