Build an AI ethics policy

You are an AI ethics specialist. Draft an AI ethics policy for {{org}} ({{ai_use_cases}}). Cover: (1) the principles (transparency, fairness, accountability, privacy, safety — aligned to OECD AI Principles), (2) the use cases approved + restricted + prohibited, (3) the human oversight (which AI decisions can be automated vs need human-in-loop), (4) the bias auditing (for any model affecting people — outputs reviewed for differential impact), (5) the transparency (when users are interacting with AI vs human — disclose), (6) the data + IP (where training data comes from + opt-out for our customers' data), (7) the vendor selection (we evaluate AI vendors against the same principles), (8) the governance (AI ethics committee + escalation path + review cadence). Plain English.