OpenAI Faces 30 New Lawsuits Over ChatGPT’s Alleged Role in Tumbler Ridge School Shooting

OpenAI and CEO Sam Altman are facing 30 new federal lawsuits from students, teachers and a school principal affected by the Tumbler Ridge mass shooting, expanding litigation that claims the company failed to alert Canadian authorities after its safety team identified the shooter as a potential threat months before the attack.

The complaints, filed Wednesday in the Northern District of California, bring the number of lawsuits tied to the February shooting to 37. The latest cases broaden the group seeking damages to include students and educators who were inside Tumbler Ridge Secondary School during the attack, along with the family of a student who was shot.

At the center of the litigation is OpenAI’s handling of ChatGPT activity linked to 18-year-old Jesse Van Rootselaar. The lawsuits allege the company flagged conversations involving gun violence in June 2025 and referred the account for human review.

Members of OpenAI’s safety operation allegedly determined Van Rootselaar presented a credible threat and recommended contacting the Royal Canadian Mounted Police. The company instead deactivated the account without notifying Canadian authorities.

The new complaints provide more detail about who allegedly made that decision. They claim OpenAI’s Intelligence and Investigations team recommended contacting police but was overruled by company leadership.

The filings link the decision to the chain of command involving Chief Global Affairs Officer Chris Lehane and Altman. They allege Lehane either made the decision or had authority over those who did and that Altman approved or ratified it.

OpenAI disputes those allegations. Chief Strategy Officer Jason Kwon has said Lehane was not involved in the original referral decision and that investigators did not report to him. The company also denies that political or public relations concerns influenced the decision.

OpenAI has said Van Rootselaar’s activity did not meet its threshold at the time for an imminent and credible risk of serious physical harm that warranted contacting law enforcement. After the first account was deactivated, Van Rootselaar created another account and continued using ChatGPT, according to the lawsuits. OpenAI says it did not know about the second account until after the shooting.

The company’s safety policies allow conversations to be referred to law enforcement when reviewers determine there is an imminent and credible risk of serious harm to others. Altman later apologized to the Tumbler Ridge community for OpenAI’s decision not to contact Canadian police and said the company was working to strengthen its safety procedures.

Van Rootselaar killed her mother and 11-year-old brother at their home on Feb. 10 before going to Tumbler Ridge Secondary School, where six more people were killed, and dozens were wounded. Van Rootselaar died by suicide during the attack.

The lawsuits accuse OpenAI and Altman of negligence, aiding and abetting, and product liability, among other claims. A central legal question is whether OpenAI can be held responsible for harm committed by a user after the company allegedly identified a risk but did not alert authorities.

Negligence generally requires showing that someone had a legal duty to use reasonable care, failed to meet that duty, and caused harm. That duty can be harder to establish when the harm was directly caused by someone other than the company being sued.

The Tumbler Ridge cases rely in part on a doctrine known as negligent undertaking. Under California law, a company that voluntarily takes on safety measures intended to protect others can sometimes face liability if it performs that work without reasonable care and the failure increases the risk of harm or causes others to rely on those protections.

The families and survivors argue OpenAI assumed such a responsibility by building systems to identify dangerous activity, reviewing Van Rootselaar’s account and deciding whether a threat should be reported to police. OpenAI can challenge whether those measures created a legal duty and whether its actions caused the injuries alleged.

Aiding and abetting carries a higher legal hurdle. California courts generally require knowledge of the wrongful conduct and substantial assistance or encouragement in carrying it out, not simply evidence that someone could have done more to prevent the harm.

The lawsuits claim OpenAI crossed that line based on what it allegedly knew about Van Rootselaar’s conduct and the role ChatGPT played before the shooting. The company can dispute both the extent of its knowledge and whether its conduct amounted to the assistance required to establish the claim.

The cases also assert product liability claims, alleging flaws in ChatGPT’s design and safety warnings. Those claims ask whether traditional rules governing defective products can apply to an artificial intelligence service.

OpenAI has separately asked the federal court to dismiss the earlier Tumbler Ridge cases so the litigation can proceed in British Columbia, arguing that many witnesses, records and other evidence are located in Canada. The 37 lawsuits remain pending in federal court in Northern California.