Fraud Detection

Fraud Detection from ComplyAdvantage stops this fraud using a combination of:

• ML Anomaly Detection Algorithms
• Behavioral Analytics
• Time Sequencing
• NLP Neural Networks (Ex. Transformer that’s also used in ChatGPT)

It detects:

1. Changes in behavior, activity from strange and unseen counterparties, and money transfers to high-risk countries or banks
2. Changes in behavior following a suspicious event, such as a failed password attempt or a change in address
3. Activities at uncommon times to a particular user (such as a user transacting at 4 am)
4. Discrepancies between payments reference texts & payment details

Read more about ATO fraud detection in our blog.

Fraud Detection from ComplyAdvantage stops this fraud using a combination of:

• ML
• Behavioral Analytics
• Time Sequencing
• Graph Network Analysis

It detects:

1. Anomalous behavior indicating a vulnerable person has been taken advantage of by a third party
2. Discrepancies between payment reference texts & payment details
3. Changes in periodic behavior indicative of invoice fraud, such as a change in a bank account number on a typical payment
4. Synthetic identities or mule accounts used to hide the destination of the funds

Read more about payment fraud detection in our blog.

Fraud Detection from ComplyAdvantage detects synthetic identity fraud by applying a combination of:

• Identity Clustering
• Other ML algorithms

It identifies networks of accounts that are acting in tandem, both from a behavioral and a personal data perspective. It gives you dual protection, catching both:

• Accounts that look similar data-wise but act differently
• Accounts that act similarly but have different sets of personal information

Read more about payment fraud detection in our blog.

Fraud Detection from ComplyAdvantage stops this fraud using a combination of:

• Behavioral Analytics
• Identity Clustering
• Peer Group Analysis
• K-means Clustering

Using these techniques, our solution can detect:

1. Merchants fraudulently representing themselves as another type of business - for example, a casino claiming to be a restaurant
2. Merchants with significant refunds and chargebacks compared to others or previous activity from the same merchant
3. Businesses with similar patterns to others you have offboarded
4. Companies taking lines of credit with an anomalous “bust-out” pattern

Read more about payment fraud detection in our blog.

Fraud Detection from ComplyAdvantage stops this fraud using a combination of:

• Behavioral Analytics
• Time-based Aggregations
• ML Anomaly Detection

Using these techniques, our solution can detect:

1. Customers that process large amounts of refunds or chargebacks, either at a total level or in comparison to their peer group
2. Anomalous patterns of refunds or chargebacks indicative of fraud

Find out more about chargeback fraud detection in our blog.

With our pre-built scenarios and ML techniques, our solution can detect individual cases of credit card fraud, and identify networks of stolen information across accounts that may slip through traditional systems.

• Behavioral Analytics
• ML Anomaly Detection
• Identity Clustering
• K-means Clustering

Using these techniques, our solution can detect:

1. Test transactions followed by anomalous activity for a customer or their peer group
2. Overall changes in behavior by customers indicating fraud
3. Clustered networks of stolen information used in tandem by fraudsters through similar behavior and details

Read more about payment fraud detection in our blog.