Found an error?
Report now
ARCHIVE - OpenAI CEO Sam Altman during an appearance in Berlin in February 2025. photo: Sebastian Christoph Gollnow/dpa/Archivbild
Keystone
Two and a half months after the deadly shooting at a school in Canada, OpenAI CEO Sam Altman has apologized for the fact that the ChatGPT operator had not informed security authorities about suspicious AI chats of the alleged future perpetrator. "I am deeply sorry for this," Altman wrote in a letter to the residents of the town of Tumbler Ridge.
The 18-year-old suspect is alleged to have killed eight people, six of them in a school, in the western Canadian town of Tumbler Ridge on February 10. Among the victims were students, a teacher and the mother and stepbrother of the suspected shooter, who then killed herself.
OpenAI had blocked an account of the alleged perpetrator last June after automated monitoring systems noticed scenarios involving gun violence in her conversations. However, the company did not consider it necessary to notify the authorities. After the crime, OpenAI discovered another ChatGPT account belonging to the alleged shooter.
Altman assured in the letter that OpenAI wanted to find ways to prevent "such tragedies" in the future. The letter was published by the local news website "Tumbler RidgeLines", OpenAI confirmed the authenticity of the document.
According to police, the suspect was known to have mental health issues. According to investigators, she had been registered as male at birth. The transition to female had begun years before the crime, they said shortly after the fatal shooting.