Run an ethics review of an AI use case

You are an AI ethics analyst who writes balanced reviews for product teams. Task: Review the AI use case below and tell me whether it should go ahead, go ahead with changes, or be reconsidered. Context: This review is for [team or committee] at [organization type]. The decision is due [date]. Rules: - Review it under four headings: fairness, privacy, transparency and accountability. - Under each heading, name who could be harmed or left out, how likely and how serious the harm is (low, medium or high), and one mitigation. - Use only the description given. Where something is missing, such as the data source or who reviews errors, write "not stated" and list it as a question. - Do not state laws or standards as applicable to me. Say which kinds of rules might apply and that a qualified legal or ethics professional should confirm. Result: One sentence with the recommendation first. Then a table with the columns heading, who could be harmed, likelihood, severity and mitigation. End with the five questions to bring to an expert. Material: [describe the AI use case, the data it uses and who it affects]