OpenAI has refused to share the Tumbler Ridge mass shooter’s chatbot history with the B.C. government, Attorney General Niki Sharma said Monday while announcing the province’s lawsuit against the tech company has been filed in California. Sharma said the lawsuit, which was first announced back in July, will focus on OpenAI’s “failure to notify law enforcement” of shooter Jesse Van Rootselaar’s concerning ChatGPT use prior to February’s devastating attack. “The conversation between ChatGPT and the shooter was so alarming that it triggered an internal review by OpenAI employees,” Sharma told reporters at a news conference in downtown Vancouver. “We owe it to the families, survivors and everyone whose life was changed by this tragedy to get answers and accountability.” The artificial intelligence giant has previously acknowledged Van Rootselaar’s ChatGPT communications were concerning enough for her account to be banned. That ban was bypassed when the shooter opened a second account. The details of what was said—and how the chatbot responded—have never been revealed. Sharma confirmed the government has asked for the records to be disclosed, but was refused. “We should all be asking them why,” she added. Similar lawsuits have already been filed in California, where OpenAI is headquartered, on behalf of the victims and families of the Tumbler Ridge tragedy. Last week, OpenAI filed motions to dismiss dozens of the claims, arguing the matter should be tried in B.C. instead because the shooting “implicates British Columbian institutions, law, and public policy, including local law enforcement, mental-health treatment, firearms regulation, and schools.” The company also said because many witnesses and records are located in B.C., evidence can be more easily compelled if the lawsuits proceed in the province. Asked why the government chose to file its lawsuit in the U.S. as well, Sharma said the decision not to report Van Rootselaar’s ChatGPT use happened south of the border. None of the allegations against the company have been proven or tested in court. The latest lawsuit comes amid heightened scrutiny of artificial intelligence services, particularly after several major AI companies call on the U.S. government to impose stronger regulations on the industry. Sharma said she’s recommended the Canadian government consider updating the Criminal Code so there can be “human and corporate accountability for the dangerous actions of AI.” “The changes I’m asking for would ensure ai providers are placed on an equal footing with other companies and industries,” she said. “These amendments would send a clear signal to ai providers to design their products safely or face the consequences.”
