OpenAI's safety systems did exactly what they were designed to do — automated tools flagged the Tumbler Ridge shooter's account and human reviewers assessed the risk, ultimately banning her account. The failure wasn't the technology; it was a judgment call about an ambiguous threat threshold, which OpenAI has since revised. Blaming the platform ignores that multiple institutions, including law enforcement and mental health services, also missed clear warning signs.