Classify a fraud case
Answer five questions about one case to see the likely fraud type, loss bucket and next steps.
Fraud type
These choices point to third-party fraud, and the loss goes in fraud loss.
Third-party fraud is a fraud loss because the named person is a victim and the debt belongs to someone else.
Next steps
How strongly each type is indicated
Get the result by email
Put these questions into the fraud rules as a decision flow so every case gets the same label when opened, then test the rules on past cases before they go live. Start your trial
How the type is worked out
Each choice adds points to one or more fraud types. The type with the most points wins, and a tie goes to the type listed first.
| Choice | First-party fraud | Third-party fraud | Synthetic identity fraud | Money mule |
|---|---|---|---|---|
| Who opened or uses the account | ||||
| Named person says it was them | 3 | · | · | 1 |
| Named person says it was not them | · | 4 | · | · |
| No real person behind the identity | · | · | 4 | · |
| Named person let someone else use it | · | · | · | 4 |
| What identity checks showed | ||||
| Long credit history, details match | 2 | · | · | 1 |
| Real details, recent contact change | · | 3 | · | · |
| Thin file or ID used with names | · | · | 3 | 1 |
| Not checked or unknown | · | · | · | · |
| How the account paid | ||||
| No payment was ever made | 2 | 2 | 1 | · |
| Paid well, built limit, then bust-out | 2 | · | 3 | · |
| Payments are up to date | · | · | · | 2 |
| Customer disputed their own purchases | 3 | · | · | · |
| How money moved through the account | ||||
| Many payments in, sent out quickly | · | · | · | 4 |
| Credit or goods taken after account change | 1 | 2 | 1 | · |
| Ordinary spending | 1 | · | · | · |
| Links to other accounts or applications | ||||
| No links found | 1 | · | · | · |
| Shared phone, device or address across names | · | 1 | 3 | · |
| Same senders as flagged accounts | · | · | · | 3 |
| New device or login before activity | · | 3 | · | · |
The points are a simple aid for sorting one case, not a model trained on data. An analyst checks the case before closing it.
Fraud type classification
Fraud type classification sorts a case into first-party fraud, third-party fraud, synthetic identity fraud or money mule. This tool uses five choices about the account, identity checks, payments, money movement and links to other cases.
The type affects who may be owed money, which team handles the case and how the loss is recorded. Use the result as a starting point, then check the case and choose the next action.
Common questions
What are the types of fraud in lending?
This tool sorts a case into first-party fraud, third-party fraud, synthetic identity fraud or money mule. The choices about the case add points to those types.
What is the difference between first-party and third-party fraud?
In first-party fraud, the customer took credit with no intent to repay. In third-party fraud, the named person is a victim and the debt belongs to someone else.
What is synthetic identity fraud?
Synthetic identity fraud involves an identity with no real person behind it. There may be nobody real to collect from, and the loss is often written off as a credit loss.
What is a money mule account?
A money mule account moves stolen money. The case goes to the anti-money laundering team as a suspicious activity case.
How do you classify a fraud case?
Review who opened or uses the account, identity checks, payment behaviour, money movement and links to other cases. This tool adds points from those choices, and the type with the most points is the result.