OpenAI said it weighed contacting Canadian authorities months before a deadly school attack.
Its monitoring systems flagged the ChatGPT account of Jesse Van Rootselaar in June for possible violent activity.
The company reviewed whether to inform the Royal Canadian Mounted Police.
It decided the activity did not show an imminent and credible threat at the time.
OpenAI said referrals require clear evidence of planned serious harm.
Van Rootselaar later killed eight people in Tumbler Ridge, British Columbia, before taking their own life.
Police said the victims included a teaching assistant and five students aged 12 to 13.
The attacker had previous mental-health-related contact with law enforcement.
After the shooting, OpenAI voluntarily shared account information with investigators.
The company said it would continue to assist the inquiry.
The motive remains unknown.
The attack is the deadliest in Canada since the 2020 Nova Scotia killings.
