Seven lawsuits were filed in U.S. court on Wednesday against OpenAI on behalf of families impacted by the February mass shooting in the small Canadian mining town of Tumbler Ridge.
The artificial intelligence behemoth has faced intense criticism over its decision not to report the troubling ChatGPT usage of Jesse Van Rootselaar, the 18-year-old transgender woman who killed eight people at her home and a school.
OpenAI banned her account in June 2025 but said it did not report the account to Canadian police because it saw no evidence of an imminent attack.
The lawsuits filed in a U.S. federal court in California allege OpenAI decided not to report Van Rootselaar “because reporting one case would mean reporting thousands,” a statement from the legal team said.
The lawsuits also challenge the assertion that Van Rootselaar’s ChatGPT account was actually banned.
They allege that when an account is shut down for dangerous behaviour, OpenAI instructs the individual on how to resume usage, including tips on how to circumvent the 30-day suspension period.
“OpenAI also tells users that if they don’t want to wait, they can open a new account immediately using a different email address,” the statement said.
Van Rootselaar reportedly opened a second ChatGPT account after her first one was shut down.
The U.S. legal team said it is working in co-ordination with Canadian lawyers who had previously filed a lawsuit against OpenAI on behalf of the family of Maya Gebala, a 12-year-old gravely injured in the shooting.
But the U.S. actions will “supersede” the Canadian case, Wednesday’s statement said.
“There are more cases to come. Over the next several weeks, a cross-border team … will be filing over two dozen cases on behalf of the victims of the Tumbler Ridge mass shooting. The lawsuits will be filed in waves,” it added.
OpenAI CEO Sam Altman apologized to the remote community of Tumbler Ridge earlier this month, saying he “was deeply sorry that we did not alert law enforcement to the account that was banned in June.”
The company has also said that under its current security policies, which have been revised since June, Van Rootselaar’s conduct would have been flagged to police.
Asked to comment on Wednesday’s legal filing, an OpenAI spokesperson said: “We have a zero-tolerance policy for using our tools to assist in committing violence. As we shared with Canadian officials, we have already strengthened our safeguards, including improving how ChatGPT responds to signs of distress.”
Van Rootselaar killed her mother and brother at the family’s home before heading to the local secondary school, where she shot dead five children and a teacher.
She died of a self-inflicted gunshot wound after police entered the building.
TorStar headline with Minister Marc Miller proclaiming that AI was not to blame.
This ought to have been obvious to anyone that looked at the incompetence of the courts and police in their returning to guns to the troubled household.The issue however is more complex than that. I sent the following letter to the editors of 4 Canadian Newspapers shortly after the tragedy, none would publish it. You at least, had the courage to do so. My letter for your consideration follows:
”
Sam Altman is a victim of various levels of government, the law, and police, all happy to blame AI and social media for the shooting tragedy and completely absolving themselves of the responsibilities they ignored. The courts and police did not do their jobs. The shooter had been in trouble with the law multiple times and “was known to the police”. The police properly seized the guns from the home after one incident which was the correct procedure. They are allowed, by law to do so without first obtaining a warrant and they did this. Under Canadian law, seized guns must be returned to the rightful owner unless the police provide reasons to prohibit their return. There were two failures, one involving the court, the other the police. When the mother (the rightful owner) petitioned the court for the return of the guns, it would appear that the court may not have advised the police of the petition to return the weapons, or they may have advised the police who in turn did not object to the return. The law clearly states the police have this power and responsibility. We are not told what if any, communications between the BC court and police occurred. However, it was this failure to act in the interest of public safety that enabled the shooting tragedy. Neither the police nor the courts have accepted any responsibility for the return of the weapons to a troubled household.
With regards to social media being to blame there is a massive contradiction. Canada has both federal and provincial Personal Information Protection and Electronic Documents Act’s (PIPEDA). British Columbia has its own privacy act as well. These Acts are designed to protect an individual’s personal information, including medical records, from being disseminated in any way. BC’s act specifically requires that the person whose information is held by the third party, must consent to their information being disclosed. Part 3 of BC’s Act states:
Consent required
6 (1)An organization must not
(a)collect personal information about an individual,
(b)use personal information about an individual, or
(c)disclose personal information about an individual.
(2)Subsection (1) does not apply if
(a)the individual gives consent to the collection, use or disclosure,
(b)this Act authorizes the collection, use or disclosure without the consent of the individual, or
(c)this Act deems the collection, use or disclosure to be consented to by the individua
The situation is clear, on the one hand we have laws preventing the disclosure of personal information, on the other hand we have people clamoring to blame a social media site for complying with the law and not unlawfully disclosing personal information. You can’t have it both ways. Unless the law is changed, social media of all kinds is prohibited from disclosure. Imagine if this were not the case, and every online site began reporting on troublesome posts. We live in an age of deepfakes and internet bots, many fully automated and not needing humans to generate their posts. The deepfakes and bots then begin maliciously posting phony messages that are then sent to the police. These would be completely unmanageable and could easily overwhelm the police as they attempted to investigate each fake post. If we want spying on every person all the time to try to detect a problem with their behavior we have the tools. Facial recognition is becoming ubiquitous and virtually all communications can be recorded by governments and have been since before the NSA’s Echelon communications spying as far back as the 1970’s, well before the internet existed.
Governments and police have a tough virtually impossible job in trying to deal with todays plethora of social media sites and their real or fake postings. Our laws so far can’t deal with it and there is little public discussion of the conflicting needs for safety and security versus personal privacy. This needs to be discussed in public forums to define and agree modified protections and procedures. As always if new laws are proposed there needs to be a clear analysis of the problems that such new laws might generate. Tumbler Ridge was not the fault of AI or social media, it was clearly a failure of the BC courts and police. Instead, the public has been diverted to a more convenient alleged perpetrator, AI and the internet. It is time the facts were presented, without which, nothing will be solved and more tragedies can be anticipated.