Fraud Signal Lab: Spot Suspicious Behavior Without Building a Surveillance Machine

A product starts seeing suspicious refunds, invite chains, coupon abuse, or fake registrations. Participants design a fraud-response workflow that uses limited, explainable signals rather than invasive data collection or automatic punishment.

Teams will decide what to flag, what to block, what needs human review, and how users can appeal mistakes. The goal is a safety system that protects the business and community without treating everyone as a suspect.