They did the math and decided safety was an acceptable risk
OpenAI's safety team flagged the shooter's troubling ChatGPT conversations months before the attack, but leadership chose not to alert police, according to the lawsuits. The lawsuits allege OpenAI prioritized company valuation and reputation over public safety, with leadership vetoing the safety team's recommendation to report the suspect.
- Eight people killed, including six children, in Tumbler Ridge secondary school shooting on February 10
- OpenAI's safety team flagged shooter's ChatGPT conversations months before attack; leadership vetoed reporting to police
- Seven lawsuits filed in California; lawyer expects to file over two dozen more
- 12-year-old survivor Maya Gebala shot three times; remains hospitalized
- OpenAI valued at $850 billion
Seven families of victims from a February mass shooting in Canada have filed lawsuits against OpenAI and CEO Sam Altman, alleging the company ignored ChatGPT activity flagged by its safety team that referenced gun violence.
On a February morning in Tumbler Ridge, British Columbia, an 18-year-old named Jessie Van Rootselaar walked into a secondary school and opened fire. Eight people died that day—six of them children. Among the survivors was a 12-year-old girl named Maya Gebala, who was shot three times: in the head, neck, and cheek. She remains hospitalized.
What emerged in the weeks after the shooting was a sequence of decisions that now sits at the center of seven lawsuits filed this week in California. OpenAI's safety team had flagged Van Rootselaar's ChatGPT conversations months before the attack. The conversations contained references to gun violence and described scenarios involving firearms. The safety team reviewed the material, assessed the risk, and made a recommendation: report the suspect to the Royal Canadian Mounted Police. That recommendation was vetoed by executive leadership at OpenAI, according to the lawsuits filed by families of the victims and their legal team.
The families' lawyer, Jay Edelson, alleges that the company's senior leadership, including CEO Sam Altman, made a calculation. OpenAI is valued at $850 billion. The lawsuits claim that protecting that valuation and the company's reputation took precedence over alerting authorities to a potential threat. "They did the math and decided that the safety of the children of Tumbler Ridge was an acceptable risk," one lawsuit states. Edelson expects to file more than two dozen additional legal actions on behalf of victims and community members. He has requested jury trials in each case.
The lawsuits also allege that OpenAI's claim to have banned Van Rootselaar from the platform was misleading. According to the legal filings, the suspect created a new account under the same name and continued using ChatGPT to plan the attack. OpenAI has disputed this, saying it revokes access from banned users and takes steps to prevent them from creating new accounts. Edelson says he requested the suspect's chat logs from OpenAI but was refused. He expects to obtain them through discovery in the lawsuits.
Sam Altman issued an apology last week in an open letter published by a local news outlet. "I am deeply sorry that we did not alert law enforcement," he wrote. "While I know words can never be enough, I believe an apology is necessary to recognize the harm and irreversible loss your community has suffered." In the same letter, he committed OpenAI to working with government at all levels to prevent similar incidents. The company has also announced strengthened safeguards, including improved assessment and escalation procedures for potential threats of violence.
OpenAI's official response to the lawsuits emphasizes its zero-tolerance policy for tools being used to assist in violence. A company spokesperson stated that the organization has already moved to strengthen its safety measures. But Edelson's strategy is to place the jury inside the moment of decision—to show them the internal debate, the safety team members advocating for notification, and the leadership choice to remain silent. "We're going to show them how people were jumping up and down saying we need to protect this town, and we're going to show them how Sam Altman and OpenAI routinely make these decisions to put their own interests first," he told the BBC.
This is not OpenAI's only legal exposure on this front. The company is also facing a criminal probe in Florida related to ChatGPT use by a man accused of carrying out a shooting at Florida State University last year, in which two people were killed and several others injured. The pattern—AI platform, warning signs, inaction, tragedy—is beginning to repeat.
Notable Quotes
I am deeply sorry that we did not alert law enforcement. While I know words can never be enough, I believe an apology is necessary to recognize the harm and irreversible loss your community has suffered.— Sam Altman, OpenAI CEO, in open letter
We're going to show them how people were jumping up and down saying we need to protect this town, and we're going to show them how Sam Altman and OpenAI routinely make these decisions to put their own interests first.— Jay Edelson, lawyer for families
The Hearth Conversation Another angle on the story
Why would a company with that much to lose ignore its own safety team's recommendation?
The lawsuits allege it came down to risk calculation. A $850 billion valuation is fragile if the public perceives your product as a tool for violence. Reporting the threat might have drawn scrutiny, investigation, regulatory attention. Staying silent meant the risk remained theoretical—until it became real.
But they had actual knowledge of the threat, according to the filings. How do you defend that?
You don't, which is why Altman apologized. But the company's position is that they have systems in place now, that they've learned. The lawsuits are asking whether learning after the fact is enough, or whether the decision itself was negligent.
The suspect created a new account after being banned. That seems like a basic failure.
It does. OpenAI says they take steps to prevent that. The lawsuit says those steps didn't work, or weren't applied. It's a factual dispute, but it matters because it goes to whether the company's safeguards were ever real or just performative.
What does Edelson hope a jury will do?
He wants them to see the internal conversation—the safety team pushing for notification, leadership saying no. He wants them to understand that this wasn't a technical failure or an edge case. It was a choice made by people who knew the stakes.
Is there a path where OpenAI wins these cases?
Possibly. They could argue they had no legal duty to report, that the conversations were ambiguous, that they couldn't have predicted the attack with certainty. But the jury will have to decide whether a company that builds tools for communication has a responsibility when it sees someone planning violence.