AI-powered Anomaly Detection
Use AI/ML to detect abnormal events for transaction monitoring and fraud prevention.
Sumsub Anomaly Detection is designed to help you monitor transactions by detecting anomalies with remarkable accuracy. It utilizes AI and advanced machine learning algorithms to identify activities or transactions that deviate significantly from established applicant behavior patterns, indicating potential fraud.
This helps you:
- Spot unusual transaction activity at the moment the transaction is scored.
- Focus investigations on the transactions that stand out.
How Anomaly Detection works
Sumsub Anomaly Detection identifies unusual financial activities that may require immediate investigation. Our AI also provides insights into why a particular transaction was flagged. It examines various factors and highlights the specific reasons for its decision, clearly understanding potential issues.
Each finance transaction gets an anomaly score: Low, Medium, or High, depending on the severity of the behavior deviation, which is defined based on the signal types.
NoteThe recommended number of transactions for accurate detection is 25,000 of the corresponding applicant type (Individual or Company) per 30-day window.
Review Anomaly Detection results
The information about the anomaly score and the severity of the involved signals is available in the Anomaly Detector section of the target Transaction page.
The score is calculated based on the following signals.
| Type of signal | Description |
|---|---|
| Transaction frequency | How often the applicant makes transactions, compared to their history and to other applicants on the account. |
| Transaction amount | How much the applicant transfers, compared to their history and to other applicants on the account. |
| Current transaction amount anomaly | How unusual the current transaction amount is compared to the applicant’s past activity. |
| Currency usage | Which currencies the applicant uses in transactions. |
| Country activity | Which countries are involved in the applicant’s transactions, including IP-based location. |
| Device fingerprint and usage patterns | Which devices the applicant uses for transactions. |
Updated 2 days ago