The Canadian province of British Columbia has filed a lawsuit against OpenAI and its chief executive, Sam Altman, in a United States federal court in San Francisco, alleging the company failed to prevent a deadly school shooting that the province says could have been averted if OpenAI had alerted law enforcement.
British Columbia Attorney General Niki Sharma announced the legal action on Monday, September 21, according to a statement from the province and reporting from CBC News, Reuters, The New York Times and The Wall Street Journal. The case escalates a legal and political confrontation that has been building since February, when a mass shooting at a school in the rural town of Tumbler Ridge killed eight people. Per maggiori informazioni su questa storia, consulta la nostra tendenze IA.
What the Lawsuit Alleges
According to the complaint as described by multiple outlets, including The420.in and Vancouver Sun reporting on the filing, the attack took place on February 10, when 18-year-old Jesse Van Rootselaar killed her mother and stepbrother at home before driving to her former school. There, prosecutors and civil plaintiffs say, she fatally shot an educational assistant and five students aged between 12 and 13 before taking her own life.
The province's lawsuit centers on a striking allegation: that OpenAI's internal safety team flagged the shooter's ChatGPT conversations about gun violence in June 2025, eight months before the attack. Citing whistleblower accounts, the complaint states that safety personnel recommended reporting the user to police, but that senior executives overruled the recommendation. OpenAI deactivated the original account, according to the filing, but the shooter created a second account to continue using the service, and the province alleges the company took inadequate steps to stop her access or notify authorities.
"The people of Tumbler Ridge have endured an unimaginable loss, and they deserve answers and accountability," Sharma said in the statement announcing the suit, per iPhone in Canada. "Today, our government is taking the next step by filing a lawsuit against OpenAI."
Reuters reported that the province's core claim is that the shooting "could have been prevented" had OpenAI warned local law enforcement about the shooter's activity. British Columbia is seeking damages to cover emergency response costs, mental health services and the reconstruction of the targeted school, alongside court-ordered reforms to how the company monitors and reports threats of violence.
OpenAI's Response
OpenAI, in a statement carried by CBC News and attributed to spokesperson Drew Pusateri, called the Tumbler Ridge shooting an "unspeakable tragedy" and said the company remains committed to working with law enforcement and advancing its safety protocols. The company had earlier pledged stronger safety measures for Canada following previous lawsuits.
According to CTV News, Attorney General Sharma told reporters that OpenAI has refused to share the shooter's chatbot history with the provincial government, a dispute that is likely to become a central front in the litigation.
A Widening Legal Battle
The provincial case stands apart from the 37 claims already filed against OpenAI and Altman by victims' families, educators and students, according to the province's statement. Families of victims first sued earlier this year, and new lawsuits from teachers and students followed in September. The government's entry into the courtroom raises the stakes considerably: a provincial attorney general has the resources to pursue years of litigation and the standing to demand structural reforms, not just damages.
Sharma has also escalated the political dimension. In letters to federal ministers cited in the province's statement, she urged Ottawa to amend Canada's Artificial Intelligence and Data Act to set clear national safety standards for AI systems. "This case highlights the urgent need for strong national safeguards for artificial intelligence technology," she said.
Why This Case Matters for the AI Industry
La causa è tra i tentativi più aggressivi da parte di qualsiasi governo di ritenere uno sviluppatore di intelligenza artificiale finanziariamente e strutturalmente responsabile del danno presumibilmente facilitato dal suo prodotto. Gli osservatori legali esamineranno diverse questioni: se un tribunale accetta che un'azienda di intelligenza artificiale abbia il dovere di avvisare terzi sulle conversazioni degli utenti; come si comportano nella scoperta i resoconti degli informatori delle revisioni interne sulla sicurezza; e se le rivendicazioni canadesi possano essere presentate contro una società statunitense presso un tribunale di San Francisco.
Il caso arriva anche in un momento delicato per OpenAI, che ha dovuto affrontare una serie di controversie legate alla sicurezza nel 2026, comprese le segnalazioni secondo cui i suoi agenti autonomi sono sfuggiti ai test sandbox e al successivo controllo da parte dei procuratori generali dello stato. Un governo provinciale che ora sostiene che i lavoratori della sicurezza interna sono stati sopraffatti dai dirigenti aggiunge una narrazione nuova e potenzialmente dannosa a quel record.
Per quanto riguarda l’industria in generale, la denuncia mette alla prova un argomento che circola nei dibattiti sulle politiche sull’IA da due anni: che quando i chatbot assumono ruoli consultivi e di compagnia, i loro sviluppatori potrebbero acquisire responsabilità più vicine a quelle dei servizi regolamentati che delle piattaforme neutrali. Se la Columbia Britannica dovesse riuscire a stabilire anche solo una parte della sua teoria, il precedente potrebbe influenzare il modo in cui tutti i principali laboratori di intelligenza artificiale trattano gli utenti che mostrano segni di violenza.
Né OpenAI né la provincia hanno indicato quando è probabile che il tribunale ascolti le mozioni iniziali. Secondo molteplici rapporti, il caso è stato depositato presso il tribunale distrettuale degli Stati Uniti per il distretto settentrionale della California a San Francisco.
---
Rimani Aggiornato sull'IALe ultime notizie, analisi e scoperte sull'intelligenza artificiale — tutto in un posto.
Leggi altre notizie sull'IA →
