The Canadian federal government has called top safety officials from OpenAI, the U.S. company behind the ChatGPT artificial intelligence chatbot, to Ottawa to discuss why the company failed to report its safety concerns about a person who later committed a mass school shooting, Artificial Intelligence Minister Evan Solomon said Monday.
This comes after it was discovered that the internal systems of ChatGPT had flagged and banned the account of the suspected shooter months before the attack in February.
However, the company chose not to report its safety concerns to the police at the time, an incident that has raised questions in Canada about the role of tech companies in reporting safety concerns to the police.
Solomon, speaking at a news conference in the Canadian capital, said he had invited OpenAI’s best safety team from the United States to come to Ottawa on Tuesday to meet with Canadian officials.
“We will have a sit-down meeting to have an explanation of their safety protocols,” he told reporters, adding that Ottawa was considering “all options … to protect Canadians from online harm,” without specifying what legislative or regulatory measures might follow.
Read Also: House Passes Measure To Overturn Trump’s Canada Tariffs
The controversy is related to the mass shooting that occurred on February 10 in Tumbler Ridge, a small town in the northeastern part of British Columbia, where an 18-year-old named by the Royal Canadian Mounted Police (RCMP) as Jesse Van Rootselaar shot and killed eight people, including five students and an education assistant at the local secondary school, after fatally shooting family members at home, before taking his own life.
Van Rootselaar’s ChatGPT account was reported internally to OpenAI’s abuse reporting systems in June 2025 for participating in conversations that contained scenarios of gun violence and other disturbing material, according to reports from news outlets citing individuals familiar with the situation.
Employees considered whether to report the incident to law enforcement, but the company indicated it did not report the incident because it did not constitute “an imminent and credible risk” of serious physical harm. The account was later banned for violating terms of service.
OpenAI indicated in the days following the shooting that it had provided information about Van Rootselaar’s use of ChatGPT to the RCMP and would cooperate with the investigation, but the involvement of the Ottawa government indicates a rising concern among Canadian authorities about the policing of platforms by large technology companies and when they should contact law enforcement.
Solomon, who is also responsible for federal AI policy, said he was “deeply disturbed” by reports that online activity related to the shooter had not been reported to the police in a timely fashion.
“The government is in touch with OpenAI and other companies regarding their safety procedures, especially as it relates to children,” he said in a separate statement on Saturday.
“Canadians expect online platforms, including social media platforms, to have strong safety procedures in place to alert law enforcement of potential violence,” he added.
Read Also: Canadian Authorities Report 9 Dead In BC Tumbler Ridge
Solomon did not elaborate on what kind of policies might come out of the consultations, only that “all options are on the table” to deal with online harms.
There are clearly many difficult questions about user data, platform policies, and the line between privacy and public safety that will have to be answered in this case, especially in a country like Canada where privacy laws are very strict.
For OpenAI, the incident highlights the issue of how generative AI companies handle reported content and engage with law enforcement.
The company, founded in San Francisco, employs a hybrid model of automated systems and human moderation to monitor and enforce usage policy and identify instances of misuse, such as violence.
In previous comments, the company has indicated that it maintains a high threshold for reporting to law enforcement because of the potential for unnecessary reporting to raise privacy issues and distress for users and their families.
The summons issued by Canada is part of an international discussion about the use of social media and technology companies in preventing violence.
Governments in Europe and the U.S. have been struggling with the question of whether to require more stringent regulations regarding content moderation practices and the reporting of threat indicators to law enforcement, although such requirements are still under discussion.
The mass shooting in Tumbler Ridge was one of the deadliest in Canada in recent years. Unlike the United States, where mass shootings are more common, such incidents are few and far between in Canada, which has very tight gun control laws.
The incident has sparked renewed calls for greater public safety, mental health support, and the role of digital platforms in flagging concerning behavior.