OpenAI Faces 30 New Lawsuits Over Alleged Inaction Before Tumbler Ridge Shooting

When a company's own safety team raises a red flag and nobody acts on it, the people who pay for that silence are rarely the executives who made the call. They're students in a classroom. They're teachers who thought they were safe at work. That's the grim center of a new legal battle now facing OpenAI.

Thirty new lawsuits filed Wednesday in a California federal court accuse OpenAI and CEO Sam Altman of providing "substantial assistance and encouragement" to Jesse Van Rootselaar, the suspect in Canada's Tumbler Ridge school shooting. The plaintiffs are students, teachers, and the school principal who were present at the time of the attack. Their claims, first reported by TechCrunch, echo a similar set of lawsuits filed in April by victims' families.

Here's the core allegation. OpenAI's automated review system reportedly flagged conversations between the alleged shooter and ChatGPT about gun violence, well before the attack took place. Internal safety staff recommended contacting Canadian authorities. According to the lawsuits, that recommendation went nowhere. Instead, they allege, OpenAI's chief global affairs officer Chris Lehane was involved in a decision to stay quiet, out of concern for the company's "reputational and financial standing."

Think of an automated content review system like a smoke detector wired into a building's fire alarm panel. It's designed to do more than beep locally. It's supposed to trigger a response, alert the right people, and set off a chain of action that protects everyone in the building. The lawsuits argue that OpenAI's detector went off, but the alarm never reached the people who needed to hear it.

There's a second allegation that may prove just as damaging. The new filings claim that after flagging the account, OpenAI merely "deactivated" it rather than issuing a system-wide ban. If true, that distinction matters. Deactivation can be a soft, reversible measure, closer to a temporary timeout than a permanent lock. A system-wide ban is meant to prevent someone from simply creating a new account and picking up where they left off. The lawsuits suggest the softer approach left a door open that a full ban would have closed.

A Pattern Beyond One Case

This isn't an isolated dispute. OpenAI is also named in a lawsuit brought by the state of Florida, which accuses the company of aiding and abetting mass shooters, including the suspect behind last year's attack at Florida State University. Taken together, these cases are starting to sketch a pattern that regulators, courts, and the public will need to reckon with: what legal responsibility, if any, does an AI company bear when its product is present in the lead-up to violence, and its own systems noticed something troubling before it happened?

This is genuinely new legal territory. Courts have long grappled with how much responsibility platforms bear for what users do with their tools, from social media incitement cases to gun manufacturer liability. But "aiding and abetting" is a serious legal standard, typically requiring more than passive hosting. It usually implies some form of knowing or reckless facilitation. The Tumbler Ridge lawsuits are trying to clear that bar by pointing directly at an internal recommendation that authorities be notified, and the alleged decision not to follow it.

OpenAI has pushed back hard. Jason Kwon, the company's chief strategy officer, responded on X, calling the lawsuits' claims about its safety teams "false." He said it is "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." That's a direct denial of the central allegation, that business concerns outweighed a safety recommendation. The company hasn't offered a detailed account of what actually happened internally, at least not one included in the reporting so far, which leaves the two sides at a genuine factual standoff heading into litigation.

It's worth sitting with why these particular details, the flagged conversations, the internal recommendation, the deactivation instead of a ban, carry so much weight. In most AI harm cases, the hardest part for plaintiffs is proving the company knew, or should have known, that something dangerous was unfolding. Here, the lawsuits claim there's a paper trail: a safety team that did its job, and a corporate response that allegedly didn't match the warning. If that's accurate, it shifts the legal conversation from "could OpenAI have known" to "what did OpenAI do once it knew."

None of this has been proven in court. These are allegations in a civil complaint, and OpenAI disputes the characterization of its internal process. But the fact that we're now looking at 30 new plaintiffs, on top of the families who sued in April and a state attorney general's office in Florida, suggests this isn't a fringe legal theory. It's becoming a coordinated push to establish that AI companies can be held liable not just for what their models generate, but for what their companies choose to do, or not do, once warned.

Why It Matters

This case sits at the intersection of two things society has struggled to govern well: gun violence and rapidly deployed AI systems. Families and survivors are asking a straightforward question with enormous stakes. If a company builds a system capable of detecting danger signs, does it have an obligation to act on what that system finds, even when acting might be costly or embarrassing? The answer will shape not just OpenAI's future, but the standards every AI company operating public-facing tools will eventually be judged against. For the students and teachers who lived through Tumbler Ridge, that answer can't come soon enough.