OpenAI Sued for Not Reporting a School Shooter’s ChatGPT Chats to Police

Seven families from Tumbler Ridge, British Columbia, have filed lawsuits against OpenAI and CEO Sam Altman. They claim the company was negligent for not alerting police to the suspected shooter’s ChatGPT activity before the attack.

The suspect, 18-year-old Jesse Van Rootselaar, had reportedly been having conversations with ChatGPT about gun violence. OpenAI’s own systems flagged this activity, according to the Wall Street Journal. The company “considered” reporting it to law enforcement but ultimately stayed quiet.

The families’ argument is blunt: OpenAI knew something was off, chose to protect its reputation and upcoming IPO over public safety, and people died or were injured as a result. That’s a heavy accusation, and it’s not going to be easy to prove in court. But it raises a question that’s been hanging over the AI industry for a while now: when does a chatbot’s awareness of harmful intent trigger a real-world obligation to act?

OpenAI has safety systems that flag certain types of content, including discussions around self-harm, violence, and planning attacks. These systems work reasonably well at the technical level. What they don’t do well is decide what to do with the information after the flag. In this case, it seems the company made a calculated business decision. I don’t envy that position, but I also don’t think it’s defensible when lives are at stake.

The lawsuit argues that OpenAI had a duty to report credible threats. The company’s response will likely lean on terms of service and the fact that they’re not law enforcement. That’s technically true, but it feels thin when you’re sitting across from families who lost children.

What bothers me most here is the pattern. This isn’t the first time an AI company has known about potentially dangerous user behavior and done nothing meaningful. The industry loves to talk about safety in press releases and white papers. When the rubber meets the road, the priority is always the bottom line.

OpenAI is heading toward an IPO. Lawsuits like this are bad for business. But the families aren’t asking for money they can use to replace their kids. They’re asking for accountability. And honestly, they deserve it.

This case will set a precedent. If OpenAI loses, every AI company with a content moderation system will have to rethink how they handle threat indicators. If they win, we’re basically saying that knowing about a potential attack and doing nothing is acceptable corporate behavior. Neither outcome is comfortable.

I’ll be watching this one closely. The legal arguments aside, the moral calculus here is ugly, and no amount of IPO money changes that.