In a letter posted Friday, OpenAI CEO Sam Altman apologized to the Tumbler Ridge community for that assessment. "I am deeply sorry that we did not alert law enforcement to the account that was banned in June," Altman wrote to the British Columbia town still grieving from February's shooting.

The apology reveals how OpenAI's internal threshold policy created a gap between AI detection and police intervention. While OpenAI's algorithms successfully identified concerning content in June, its review process concluded Van Rootselaar's account "didn't meet a threshold for referral to law enforcement," according to The Guardian.

OpenAI banned Van Rootselaar's account in June for policy violations but applied their threshold criteria and decided against police notification. Only after the February 2026 massacre did the company disclose it had previously flagged the shooter's account to authorities.

British Columbia Premier David Eby said it "looks like" OpenAI had the opportunity to prevent the shooting. In his response to Altman's apology, Eby called it "necessary, and yet grossly insufficient for the devastation done to the families of Tumbler Ridge."

The case exposes questions about how OpenAI establishes and applies its law enforcement referral criteria. OpenAI's systems can identify concerning content with increasing sophistication, but the company maintains internal thresholds that must be met before alerting police—thresholds that in this case weren't triggered despite the violent content detection.

Altman's letter, dated April 23 and published in the local newspaper Tumbler RidgeLines, acknowledged conversations with Tumbler Ridge Mayor Darryl Krakowka and Premier Eby. He said community leaders "conveyed the anger, sadness and concern" residents felt about OpenAI's threshold assessment.

The CEO delayed his public apology for months, writing that "time was also needed to respect the community as you grieved." He committed to "working with all levels of government to help ensure something like this never happens again."

The incident raises questions about how OpenAI establishes these referral thresholds and whether current criteria adequately balance user privacy with public safety. Unlike social media platforms that primarily host user-generated content, ChatGPT users engage directly with OpenAI's AI systems, potentially creating different considerations for intervention protocols.

OpenAI has not disclosed what specific content triggered the violence flag or detailed the threshold criteria that prevented police notification. The San Francisco technology firm said it "considered whether to refer the account to the Royal Canadian Mounted Police" but applied its established standards for such referrals.

For Tumbler Ridge, population roughly 2,000, the revelation adds another layer of complexity to an already devastating loss. Eight people died across two locations in what became one of Canada's deadliest school shootings, and now residents must grapple with knowing a tech firm had advance warning but applied policies that kept authorities uninformed.

Altman closed his letter acknowledging the permanence of the community's loss: "I cannot imagine anything worse in this world than losing a child. My heart remains with the victims."