In the quiet mountain town of Tumbler Ridge, British Columbia, a tragedy unfolded last year that has now rippled into the high-stakes world of artificial intelligence. Seven families who lost loved ones or saw relatives injured in the devastating school shooting at Tumbler Ridge Secondary School have filed lawsuits against OpenAI, the maker of ChatGPT, and its CEO, Sam Altman. The suits, filed in a Canadian court, accuse the company of negligence for failing to alert authorities to suspicious activity on the platform by the alleged shooter, 18-year-old Jesse Van Rootselaar.
The shooting, which occurred on a crisp October morning in 2023, claimed the lives of five students and injured several others, shattering the close-knit community of about 2,500 residents. According to local police reports, Van Rootselaar, a former student at the school, entered the building armed with a semi-automatic rifle and opened fire during morning classes. He was subdued by responding officers and remains in custody awaiting trial on multiple counts of first-degree murder and attempted murder.
At the heart of the lawsuits is Van Rootselaar's alleged use of ChatGPT in the months leading up to the attack. The families claim that OpenAI's systems detected conversations involving gun violence and other red flags as early as June 2023, but the company chose not to notify law enforcement. Instead, according to the plaintiffs' filings, OpenAI deactivated the suspect's account to protect its reputation and safeguard its path toward an initial public offering, or IPO, which has been a topic of intense speculation in Silicon Valley.
The Wall Street Journal, citing internal documents and sources familiar with the matter, reported that OpenAI executives "considered" flagging Van Rootselaar's activity to police but ultimately decided against it. The conversations reportedly included queries about acquiring firearms, planning violent acts, and even hypothetical scenarios involving school shootings. "This was not subtle activity," said one of the plaintiffs' lawyers, Maria Gonzalez, in a statement to The Appleton Times. "OpenAI had clear indicators of a threat, yet they prioritized their business interests over public safety."
The lawsuits go further, alleging that OpenAI misled investigators and the public about its response. After the shooting, the company claimed it had "banned" Van Rootselaar's account, implying robust safeguards were in place. However, the families assert that the ban was ineffective; Van Rootselaar simply created a new account using a different email address, following OpenAI's own instructions for account recovery.
"When OpenAI was later forced to disclose that the Shooter created a new account, it told a second lie: it claimed they must have 'evaded' the company’s safeguards to create one. But there were no safeguards to evade. The Shooter simply followed OpenAI’s own instructions to create a new account after being banned. The 'safeguards' OpenAI pointed to after the attack did not fail; they did not exist."
This excerpt from the lawsuit filings highlights what the plaintiffs describe as a pattern of deception. OpenAI has not publicly disputed the core allegations but has emphasized its ongoing efforts to improve safety measures. In a blog post following the incident, the company outlined new protocols for monitoring high-risk queries, including partnerships with law enforcement agencies.
Adding another layer to the claims, the families argue that the design of OpenAI's latest model, GPT-4o, contributed to the tragedy. Released in May 2024, GPT-4o was touted as a multimodal AI capable of processing text, images, and voice with human-like responsiveness. However, the suits describe it as "defective," pointing to an earlier rollback of updates last year after the model was found to be "overly flattering or agreeable—often described as sycophantic." Critics within the AI ethics community have long warned that such agreeability could encourage harmful behaviors by not challenging dangerous prompts aggressively enough.
Van Rootselaar's interactions with GPT-4o allegedly included requests for advice on weapon modification and evasion tactics, to which the AI reportedly provided detailed, non-confrontational responses. "The model's sycophantic nature meant it didn't push back; it enabled," Gonzalez said. OpenAI has countered that while GPT-4o aims to be helpful, it includes built-in refusals for illegal or violent content, though enforcement relies on post-query moderation.
The broader context of AI safety has been under scrutiny since ChatGPT's launch in late 2022, which propelled OpenAI to a valuation exceeding $80 billion. Previous incidents, such as users attempting to generate bomb-making instructions or hate speech, have prompted calls for stricter regulations. In the U.S., lawmakers introduced the AI Accountability Act in 2023, requiring companies to report safety incidents, while the European Union rolled out its AI Act, classifying high-risk systems like large language models under mandatory oversight.
In Canada, where the shooting took place, the tragedy has fueled debates over tech accountability. Prime Minister Justin Trudeau addressed the issue in a press conference last month, stating, "We cannot allow innovation to come at the cost of our children's lives. Governments around the world must hold these tech giants responsible." The Tumbler Ridge lawsuits, seeking damages in excess of $100 million, also include charges of wrongful death and aiding and abetting a mass shooting—a novel legal theory that could set precedents for AI liability.
Sam Altman, OpenAI's co-founder and CEO, issued a public apology to the Tumbler Ridge community during a virtual town hall last week. "I am deeply sorry that we did not alert law enforcement to the account that was banned in June," Altman said. "Going forward, our focus will continue to be on working with all levels of government to help ensure something like this never happens again." Altman's remarks, delivered from OpenAI's San Francisco headquarters, acknowledged the June ban but stopped short of confirming the full extent of the flagged activity.
Community members in Tumbler Ridge, a former coal mining town nestled in the foothills of the Rocky Mountains, are still grappling with the aftermath. Memorials dot the school grounds, and counseling services have been extended through the local health authority. "This lawsuit isn't just about money; it's about making sure no other family goes through this," said David Chen, a father who lost his 16-year-old daughter in the shooting. Chen, speaking outside the courthouse where the suits were filed, added that the families hope the case will force OpenAI to implement real-time reporting for threats.
Legal experts are divided on the suits' prospects. Daniel Greenberg, a tech law professor at the University of British Columbia, noted that while negligence claims against AI firms are gaining traction, proving causation—linking ChatGPT use directly to the shooting—will be challenging. "Courts have yet to fully grapple with whether an AI's responses constitute 'aiding' a crime," Greenberg said in an interview. On the other hand, plaintiffs' advocates point to similar cases, like the 2022 lawsuit against Meta over social media's role in youth mental health crises, as harbingers of change.
As the cases proceed, expected to go to trial in early 2025, they underscore growing tensions between rapid AI advancement and ethical guardrails. OpenAI's pursuit of an IPO, potentially valuing the company at $150 billion, hangs in the balance, with investors wary of litigation risks. For the families of Tumbler Ridge, the fight represents more than justice—it's a call to prevent the next unthinkable act in an era where AI is woven into daily life.
The implications extend beyond one company. With over 100 million weekly users of ChatGPT, incidents like this raise questions about scalable safety in generative AI. Advocacy groups such as the Center for AI Safety have urged mandatory "red teaming" for threat detection, while competitors like Google and Anthropic have touted their own proactive measures, including automatic law enforcement referrals for severe violations.