Which service uses machine learning to detect fraudulent online activities?

• A. AWS IoT 1-Click
• B. Amazon Lex
• C. Amazon Comprehend
• D. Amazon Fraud Detector

Answer: (D)

Machine learning to detect fraudulent online activities is the key idea here. Amazon Fraud Detector is specifically built for that purpose. It lets you train ML models on historical labeled data—events that were legitimate or fraudulent—and then deploy those models to score new online actions in real time. Based on the risk score, you can automatically approve, challenge, or block requests, integrate the decision into your application flow, and monitor model performance. This direct focus on evaluating fraud risk in real time is what sets it apart from the others.

The other services serve different aims: IoT 1-Click is for provisioning devices, Lex is a conversational bot that uses ML for natural language understanding, and Comprehend analyzes text for insights like sentiment or entities. They aren’t designed to detect fraudulent online activity at scale.