Families sue OpenAI over Tumbler Ridge mass shooter’s use of ChatGPT : NPR

A woman cries at a makeshift memorial for the victims of a deadly shooting in the town of Tumbler Ridge, British Columbia. A lawsuit filed Wednesday claims OpenAI was negligent in failing to report the shooter to authorities after his account was flagged for “gun violence activity and planning.”

Paige Taylor White/AFP via Getty Images

hide caption

toggle caption

Paige Taylor White/AFP via Getty Images

The families of those injured and killed in a school shooting in Tumbler Ridge, British Columbia, are suing OpenAI for negligence in providing a dangerously defective version of ChatGPT to the shooter.

The six lawsuits, filed in federal court in San Francisco, allege that OpenAI failed to take steps that could have prevented injuries and deaths during the shooting, which took place on February 10. They claim the company failed to report the shooter’s conversations with ChatGPT to authorities, and that ChatGPT itself was a flawed product that failed to challenge the shooter or direct him to seek help in the real world.

The lawsuits are the latest aimed at holding a technology company responsible for the design of its products, a once-novel legal approach that is increasingly being used against makers of chatbots, social media and other platforms.

For those who have lost loved ones, “there’s nothing the legal system can do to make them whole again,” Edelson told NPR in an interview. He added that they hope the trial will hold OpenAI executives accountable: “They should not be trusted to have the most powerful consumer technology on the planet. »

In a statement responding to the lawsuits, OpenAI said it had a “zero tolerance” policy for the use of its tools to aid in violence:

“We have already strengthened our safeguards, including improving how ChatGPT responds to signs of distress, connecting people with local support and mental health resources,” an OpenAI spokesperson told NPR in an email.

In a lengthy blog post published Tuesday evening, OpenAI explained its policies in more detail: “When conversations indicate an imminent and credible risk of harm to others, we notify law enforcement.”

“Profit over lives”

The Tumbler Ridge shootings are among the deadliest in Canadian history. It happened when Jesse Van Rootselaar, 18, entered the local high school with a long gun and a modified handgun, authorities said. Van Rootselaar killed five students and a teacher before committing suicide. Authorities later learned she also killed her mother and 11-year-old half-brother in their home before coming to school. About two dozen other people were injured in the attack.

The lawsuits filed Wednesday allege that ChatGPT, and specifically the GPT-4o model, played a crucial role in the events at Tumbler Ridge. One of the complaints, on behalf of Maya Gebala, a 12-year-old girl seriously injured in the shooting, alleges that Van Rootselaar was on ChatGPT months before the shooting and that in June 2025, OpenAI’s automated system flagged her account for “gun violence activity and planning.”

A security team reviewed the content and urged OpenAI management to notify authorities, but the lawsuit alleges that company management instead chose to deactivate the account. They also failed to act, according to the lawsuit, when the shooter created a second account and continued his conversations with ChatGPT.

Last week, Sam Altman, CEO of OpenAI, apologized to the community:

“I am deeply sorry that we did not alert law enforcement about the account that was banned in June,” he wrote. “Moving forward, we will continue to focus on working with all levels of government to ensure that a situation like this never happens again.”

In addition to allegedly failing to notify authorities of the imminent danger, the lawsuit claims that OpenAI knowingly deployed a defective product to the public.

“The Tumbler Ridge attack was the entirely predictable result of deliberate design choices by OpenAI, made with full knowledge of the direction those choices were leading,” the complaint states. “GPT-4o was designed to accept, reinforce, and expand on users’ violent thoughts rather than challenge them, interrupt them, or direct users toward real help.”

The events around Tumbler Ridge are “as clear a demonstration as possible of the moral hazard that comes with centralizing security authority in a place like OpenAI,” said Tim Marple, who worked at OpenAI in the threat-scouting division. Marple, now co-director of Maiden Labs, a nonprofit that works to identify AI risks, said he was not surprised the company failed to contact authorities.

“When I worked there and since I left, the only things I see characterizing their behavior are incompetence and greed,” said Marple, who is not associated with the latest lawsuit. He believes regulation, including mandatory reporting laws, is needed to prevent similar tragedies from happening again.

Free speech concerns

But not everyone agrees that lawsuits and regulation will help prevent tragedies like the one at Tumbler Ridge.

“What drives someone to commit an atrocity is often unclear,” said Eric Goldman, associate dean for research at Santa Clara University School of Law. Goldman worries that too strict regulation could make chatbots less useful to those who need them. He also rejects the idea that chatbots should be treated as defective products. For him, the issue is truly a question of freedom of expression.

“I would ask very difficult questions about a trial like this. Is this really the right way to regulate speech, even though in some cases speech can contribute to people making bad choices in their lives?” Goldman said.

Regardless, Goldman said complaints of negligence and defective products are on the rise. “These legal theories constitute the new frontier of Internet law,” he said.

More to come

The number of civil and criminal investigations into AI companies is increasing, agrees Meetali Jain, executive director of Tech Justice Law, an advocacy group critical of the tech industry that has been involved in several lawsuits against major companies.

Jain’s group helped represent the family of a teenager who died by suicide after having lengthy conversations with a chatbot created by the company Character.AI. That case is currently in settlement talks, but she’s hearing more and more examples of AI chatbots causing problems: “Last year, we started getting stories from people who were hurt” by bots from many different companies, she said.

Jain said she expects to see even more lawsuits like those filed Wednesday in the future. In the absence of strict regulation, Jain said civil lawsuits provide “a bulwark against AI companies that continue to act recklessly and without restraint.”