OpenAI Chief Executive Sam Altman apologized to a Canadian community for failing to alert police about a mass shooter’s conversations with its chatbot.
Authorities said Jesse Van Rootselaar, 18, killed eight people including schoolchildren in Tumbler Ridge, British Columbia, before taking her own life in February.
“I am deeply sorry that we did not alert law enforcement to the account that was banned in June,” Altman said in a letter Thursday. “While I know words can never be enough, I believe an apology is necessary to recognize the harm and irreversible loss your community has suffered.”
On Friday, British Columbia Premier David Eby posted the letter on social media.
The letter came after the Wall Street Journal, citing people familiar with the matter, reported this year that Van Rootselaar conversed with ChatGPT about gun violence, prompting OpenAI employees to debate whether to alert Canadian law enforcement.
OpenAI banned the user’s account but decided to not notify the police after looking at whether the activity would be considered an imminent and serious risk of physical harm to others.
Technology companies have faced more scrutiny in the wake of mass shootings over how criminals use their tools to plan attacks or broadcast killings. But the rise of artificial intelligence chatbots that quickly answer questions and generate content also means that people are spilling their darkest thoughts online. AI companies are now reckoning with debates about balancing public safety and privacy while also grappling with new lawsuits and investigations.
In March, the family of a Tumbler Ridge shooting victim, who was hospitalized, sued OpenAI, alleging that the company knew the shooter was planning a mass attack but failed to alert law enforcement.
In a post Friday on X, Eby called the apology “necessary, and yet grossly insufficient for the devastation done to the families of Tumbler Ridge.”
In the letter, Altman said he spoke to Tumbler Ridge Mayor Darryl Krakowka and Eby about the shooting and they agreed on a public apology. Altman said he was committed to finding ways to prevent such tragedies.
“Going forward, our focus will continue to be on working with all levels of government to help ensure something like this never happens again,” Altman said in the letter.
OpenAI is also grappling with backlash over whether it is doing enough to protect public safety in the United States.
Last week, Florida’s attorney general launched a criminal investigation into ChatGPT and OpenAI to determine whether the San Francisco AI company “bears criminal responsibility” for the chatbot’s actions in a Florida State University shooting last year that left two people dead. Prosecutors had been reviewing conversations between the suspect, Phoenix Ikner, and ChatGPT.