OpenAI and Sam Altman now face 30 additional lawsuits from students, teachers, and the principal who were at Tumbler Ridge school during the shooting. Filed Wednesday in California federal court, the suits allege the company provided “substantial assistance and encouragement” to the alleged shooter, Jesse Van Rootselaar, by failing to act after its own automated review system flagged his ChatGPT conversations about gun violence.
The core accusation is damning in its specificity. OpenAI’s safety team reportedly recommended contacting Canadian authorities, but the lawsuits claim chief global affairs officer Chris Lehane was involved in a decision to stay silent — allegedly over concerns about the company’s “reputational and financial standing.” Jason Kwon, OpenAI’s chief strategy officer, pushed back on X, calling the claims “false” and insisting that “political” or “public relations” factors never drive safety decisions. That’s a hard position to hold when internal documents, if they surface in discovery, could tell a different story.
What’s new in this wave of filings is the allegation that OpenAI merely “deactivated” the suspect’s account rather than issuing a system-wide ban. The lawsuits argue this half-measure let Van Rootselaar simply sign up again with a different email, regaining access to the chatbot. “Providing the Shooter with continued access to this tool constituted substantial assistance in itself,” the suit claims. That’s a pointed legal argument — and one that gets at a real technical weakness in how AI platforms handle banned users.
The Florida angle adds pressure. OpenAI already faces a state lawsuit there over aiding and abetting mass shooters, including the Florida State University attacker. Two jurisdictions, two different shootings, same underlying question: when does a chatbot’s output cross from neutral tool to contributing factor? Courts have been reluctant to hold platforms liable for user behavior, but the flagged-conversations detail makes this harder to dismiss than your typical Section 230 defense. If plaintiffs can prove OpenAI knew and chose silence, that’s not a moderation failure. That’s a decision.
