Reports are that Anthropic has a dedicated safety & ethics interview round. Here's how to build the first principles to crack it. Start with the distinction most candidates miss. AI safety is stopping the model from causing harm. Mechanical and observable. A guardrail that blocks medical misinformation. A confirmation gate before an agent sends an email. AI ethics is deciding what the model should and shouldn't do. Upstream and often invisible. A policy that the model won't draft termination…