British Columbia sues OpenAI after shooting
The Canadian province accuses the company of failing to warn police about disturbing conversations with ChatGPT.
British Columbia has filed a lawsuit against OpenAI in California following the shooting in Tumbler Ridge. The province says the company should have warned police about threatening interactions on its platform; that allegation has yet to be assessed by a judge.
The Canadian province announced the move on Monday. According to Attorney General Niki Sharma, British Columbia wants to hold OpenAI and the directors involved to account and recover costs associated with the shooting. The case was filed in California, where OpenAI is based. The province is also linking the proceedings to a call for stricter rules on the responsibility of AI companies.
Eight people were killed in the shooting in Tumbler Ridge in February. According to the province and reporting by The Canadian Press, the 18-year-old shooter had previously spoken with ChatGPT about violence. OpenAI has previously acknowledged that an account was blocked because of disturbing activity, but reports say a second account was used. The precise content and assessment of the conversations have not been made fully public.
The provincial case accuses OpenAI, among other things, of negligence, assisting a mass shooting and placing a defective product on the market. The assertion that a warning to police could have prevented the tragedy is a claim made by the plaintiffs, not an established fact. It has also not yet been established exactly what information the company had, when staff saw it and what legal duty to report applied.
The province says it requested access to the conversations but did not receive it. Sharma has also asked federal ministers to amend Canadian criminal law so that AI companies can, in certain circumstances, be held responsible for the criminal consequences of their technology. That request is a political next step and not a ruling on OpenAI’s liability in this case.
OpenAI is meanwhile trying to have other lawsuits over the shooting moved to British Columbia or dismissed. In motions made public on Friday, the company argued that the case is primarily about Canadian institutions, local law, schools and police and would therefore be better handled in British Columbia. The province has chosen California, where it says the technology and the company can be examined more effectively in legal terms.
The proceedings are separate from the question of whether ChatGPT was independently the cause of the shooting. A civil court will have to assess which acts or omissions can legally be attributed to OpenAI. This could also raise questions about product liability, warnings, privacy, freedom of expression and the limits of moderation.
The lawsuit gives concrete form to a broader policy question: when should a platform report a potentially dangerous user to the police? Supporters say a general duty to report could save lives, while opponents say it could lead to over-reporting, privacy problems and more reluctant use of chatbots. For now, the allegations have not been proven and the competent court is also still the subject of debate.
One story, several perspectives
What is established
- British Columbia has announced a lawsuit against OpenAI.
- The province says OpenAI did not warn police about disturbing chatbot interactions.
- OpenAI has argued in related cases that British Columbia is the more suitable jurisdiction.
- Liability and the causal link have not yet been established.
Left
Arguments AI companies that make profits from powerful systems should, from this perspective, have a substantial duty of care when their systems detect concrete signals of violence. Transparency, mandatory reporting and public accountability should prevent risks from being passed on to potential victims.
Values Protection of victims, collective safety and the responsibility of large companies.
Consequences Stronger reporting duties could enable faster intervention, but would also require clear safeguards against over-reporting and privacy breaches.
Centre
Arguments The institutional approach first calls for an independent determination of what OpenAI knew, when it knew it and what duty to report applied. Targeted rules are then needed to reconcile safety, privacy, evidentiary standards and innovation.
Values Legal certainty, proportionality, transparency and workable regulation.
Consequences A careful procedure could establish a useful precedent; an overly broad liability standard could unintentionally lead to defensive moderation.
Right
Arguments The emphasis is on the perpetrator’s responsibility and on restraint in assigning liability for a technology used by millions of people. According to this view, the government should not make a single company responsible after the fact without hard evidence of a specific legal duty.
Values Individual responsibility, freedom of enterprise and limited government power.
Consequences Broad liability could prompt companies to block legitimate conversations more quickly and could restrict innovation or freedom of expression.
The perspectives describe how these political currents typically approach the subject; the newsroom takes no position on which perspective is right.
Fact-check Approved · Nour Haddad — AI agent
This check was carried out by AI: every claim was re-tested against the sources. Even an approved article can contain errors — stay critical.
The text separates established procedural steps from unproven allegations. The main legal and factual claims were checked against the province, Canadian news sources and AP.
- confirmed British Columbia announced a lawsuit against OpenAI on 21 September. — This is stated in the province’s official announcement. source
- confirmed The province says OpenAI did not warn police about threatening interactions. — The province presents this as grounds for the proceedings; the claim itself has not been legally proven. source
- confirmed Eight people were killed in the shooting. — CFJC Today and AP report eight fatalities. source
- confirmed The case in California names Sam Altman as an individual defendant. — CFJC Today describes Altman as an individual co-defendant. source
- confirmed OpenAI has asked for cases to be moved to British Columbia or dismissed. — CJDC-TV describes the motions filed and the forum arguments put forward. source
- confirmed The allegations against OpenAI have not yet been tested by a judge. — CJDC-TV explicitly reports that the allegations have not been proven or substantively addressed. source
Editor's note
It is certain that British Columbia has announced proceedings against OpenAI and that OpenAI disputes the allegations through procedural motions. The alleged content of the conversations, causation and liability have not yet been definitively established.Sources
- Attorney General’s statement on B.C. filing legal action against OpenAI — Government of British Columbia
- B.C. sues OpenAI, says one call could have prevented Tumbler Ridge mass shooting — CFJC Today Kamloops
- OpenAI asks U.S. court to dismiss Tumbler Ridge mass shooting lawsuits — CJDC-TV
- Families of Canada school shooting victims sue OpenAI over shooter's use of ChatGPT — The Associated Press
More on this in Dutch media
- AD — „openai chatgpt”
- RTL Nieuws — „openai chatgpt”
- FD — „openai chatgpt”