Question

A streaming application ingests customer events and forwards them through a message broker to multiple downstream consumers (analytics, fraud detection, and a data lake). A privacy engineer discovers that raw event payloads containing email addresses and device IDs are being written to broker topics before any masking occurs, and each consumer applies its own inconsistent redaction logic. Which design change BEST enforces privacy-by-default across all consumers with the least duplication of controls?

Accepted Answer

Insert a stream-processing transformation stage that tokenizes the identifiers before events are published to the broker topics

Answer

Require each downstream consumer to standardize on the same redaction library and document it in a shared runbook

Answer

Encrypt the broker topics at rest so the raw identifiers are protected while stored on the broker

Answer

Add a data loss prevention scanner that alerts the privacy team whenever unmasked identifiers appear in a consumer's output