The shooting was carried out by 18-year-old Jesse Van Rootselaar, who authorities said killed her mother and 11-year-old half-brother before going to the school. She then killed five students and a teacher, injured around two dozen others and later killed herself.
OpenAI had identified concerning activity on Van Rootselaar’s ChatGPT account months earlier. According to the lawsuits, the company’s automated systems flagged the account in June 2025 for activity involving gun violence and planning. OpenAI subsequently disabled the account, but Van Rootselaar opened another account and continued using ChatGPT. The company has said it did not know about the second account until after the shooting.
The new complaints focus more closely on what happened inside OpenAI after the original account was flagged. Plaintiffs allege members of the company’s Intelligence and Investigations team reviewed Van Rootselaar’s conversations, concluded they represented a credible threat of violence against real people and recommended notifying the Royal Canadian Mounted Police.
According to the complaints, that recommendation was ultimately rejected. The plaintiffs allege OpenAI Chief Global Affairs Officer Chris Lehane, or someone reporting through his organization, played a role in the decision and that Altman approved it.
One complaint filed by a teacher at the school alleges that the final decision was taken out of the hands of OpenAI personnel trained to assess threats and instead made higher in the company’s management structure.
Jay Edelson, the lead attorney representing the plaintiffs, said the cases are intended to examine how those decisions were made inside OpenAI. “What these complaints show is the larger impact and how many people's lives were really destroyed all based on decisions made by people at OpenAI,” Edelson said.
OpenAI strongly disputes the allegations involving Lehane and the company’s decision-making process. “It is absolutely false to say Chris Lehane was involved with our original referral decision, or that our investigators report to him in any way,” OpenAI Chief Strategy Officer Jason Kwon said.
Kwon also rejected the allegation that political or communications concerns influenced the company’s safety decisions. He said OpenAI’s investigations team and legal department report to him and are responsible for developing and applying standards for when cases should be referred to law enforcement. “It’s also completely untrue to say that the people at the center of these challenging decisions do not prioritize safety, or that there are ‘political’ or ‘public relations’ factors at play,” Kwon said.
The complaints do not provide direct evidence that Lehane personally overruled the safety team, and Lehane is not named as a defendant. Edelson said the legal team is relying partly on its understanding of OpenAI’s organizational structure as well as information gathered through its broader investigation. He said Lehane, Altman and members of the safety team are expected to be important witnesses.
OpenAI has previously said Van Rootselaar’s activity did not cross its threshold for an imminent and credible threat that would justify contacting law enforcement. Kwon said the company must weigh indications of potential violence against user privacy when making those decisions. “That judgment is not infallible, but it is always rooted in looking out for this balance for people,” Kwon said.
The new lawsuits challenge that explanation by pointing to a separate incident in November 2025, when OpenAI restricted access to its San Francisco offices after an alleged threat involving an activist. The complaints argue that the company notified police and warned employees despite acknowledging there was no indication of an active threat.
The plaintiffs contend that the different responses show OpenAI was willing to act before a threat became imminent when its own employees were potentially at risk.
OpenAI has said it maintains a “zero tolerance” policy against using its products to facilitate violence. Following the first Tumbler Ridge lawsuits, the company said it had strengthened its safeguards and its handling of conversations involving distress or threats. Its policy states that law enforcement is contacted when conversations indicate an imminent and credible risk of harm to others.
Altman also apologized to the Tumbler Ridge community for OpenAI’s failure to notify law enforcement after the company initially flagged Van Rootselaar’s account.
The Tumbler Ridge litigation adds to a series of legal challenges involving alleged harms connected to AI chatbots. More than 30 lawsuits had already been filed against OpenAI and other AI companies before the latest complaints, involving suicides, shootings and other alleged mental or physical harms.
OpenAI is also facing a lawsuit from the state of Florida accusing the company and Altman of failing to adequately warn users about potential dangers associated with ChatGPT while marketing the service as safe and reliable, including for children. Florida authorities are separately investigating OpenAI in connection with a shooting at Florida State University in which the accused gunman allegedly used ChatGPT for advice.
The latest Tumbler Ridge cases could put greater scrutiny on how OpenAI decides when threatening user activity should be reported to authorities, as well as who inside the company has final authority over those calls.
This analysis is based on reporting from NPR & TechCrunch.
Image courtesy of Christinne Muschi/The Canadian Press via AP.
This article was generated with AI assistance and reviewed for accuracy and quality.
