Skip to main content

Fraud Detection and Prevention

  • Anomaly Detection: Identify unusual patterns in transaction data that may indicate fraudulent activity.
  • Real-time Fraud Detection: Use real-time analytics to detect and prevent fraudulent transactions as they occur.
  • Behavioral Biometrics: Analyze user behavior (e.g., face detection, liveness in KYC) to detect potential fraud.

Amazon Fraud Detector (Deprecated)​

Amazon Fraud Detector availability change - Amazon Fraud Detector

  • Thank you for your interest in Amazon Fraud Detector. After careful consideration, we have made the decision to no longer accept new customers as of November 7th, 2025.
  • If you're looking for a fraud detection solution, we recommend AutoGluon, which is an open-source automated machine learning (AutoML) library. More details are available at the AutoGluon website and the AWS Open Source Blog. The AutoGluon Fraud Detection notebook can be found here on Kaggle. A general framework notebook is here for Amazon SageMaker AI notebook. After training AutoGluon models, you can use SageMaker AI for deploying models (more information here). AWS also has a workshop built to help you set up real-time payment processing architecture.

Amazon Fraud Detector is a fully managed service that makes it easy to identify potentially fraudulent online activities such as online payment fraud and the creation of fake accounts.

Amazon Fraud Detector

  1. Step 1 - Explore data models for your business use case.
  2. Step 2 - Define the event you want to evaluate for fraud.
  3. Step 3 - Upload your historical event dataset to Amazon S3 or stream and store your event data directly in AFD.
  4. Step 4 - Select a model type and train your model. The service automatically inspects and enriches data, performs feature engineering, selects algorithms, trains and tunes your model, and hosts the model.
  5. Step 5 - Create rules to either accept, review, or collect more information based on model predictions.
  6. Step 6 - Call the Amazon Fraud Detector API from your online application to receive real-time fraud predictions and take action based on your configured detection rules.

Models​

  • Transaction Fraud Insights
  • Online Fraud Insights
  • Account Takeover Insights

Model metrics​

Model performance charts

ROC Curve

Model variable importance

Others​

  • Nametag: Identity Verification & Account Protection Solutions
    • Prevent breaches and reduce IT support costs with ready-to-use solutions for secure onboarding, self-service account recovery, helpdesk verification and agentic AI security. Built on Deepfake Defense™ identity verification and integrations with your identity stack.
    • Data breaches
    • Ransomware
    • Social engineering
    • AI deepfakes
    • Fake IT workers
    • Account takeovers
    • Insider threats
    • Presentation attacks
    • Injection attacks