"Routed to a human" is what the suicide hotline numbers do. OpenAI employees are neither trained nor credible to do that stuff.