AI doesn't remove bias — it scales it. That makes responsible AI a leadership issue, not a technical one.