What Are Synthetic Identities?
Synthetic identities are a combination of first name, last name, birth date, and social security number that have a credit history but aren’t real people.
How is this possible, you ask?
When credit bureaus receive applications, they use these key pieces of information to match an individual to their credit history. Because the Social Security Administration doesn’t share ownership of social security numbers with them, the credit bureaus can’t know if a person uses their correct information on a loan application. When credit bureaus see a new combination of identity information that they can’t match to an existing credit history, they create a new one.
Fraudsters know this.
Who Creates Synthetic Identities?
While fraudsters are a major source of synthetic identities, not all synthetic identities are malicious. New immigrants sometimes create synthetic identities because many lenders won’t lend to them with their ITIN. Individuals looking to improve their credit may unknowingly create synthetic identities when credit repair agencies advise them to use “Credit Privacy Numbers” (social security numbers that don’t belong to them) to protect their identity. In many cases, these individuals do get a loan and go on to successfully repay the loan.
When that’s not the case, however, lenders’ losses are severe. More than $1 billion in US auto fraud due to synthetic identities occurs annually. The problem is the combination of real and fake identity data. It makes it challenging to differentiate a synthetic identity from a paying synthetic identity or a true identity.
Synthetic ID Red Flags
Synthetic identities have many commonalities that make them different from true identities, including:
- Time on File is generally short, especially for the age of the individual
- Authorized Tradelines are often present and drive the credit score upwards
- Fake Tradelines or current tradelines reported by now-defunct companies may be present
- Multiple Identities are often associated with the same SSN
- Deceased SSNs or SSNs issued long before the individual was born often appear
But some of these characteristics also appear in legitimate credit profiles, which further exasperates detection efforts.
How Does Point Predictive’s FraudBot Catch Synthetic ID Fraud?
Because Point Predictive’s FraudBot can quickly scan thousands of loan applications across many lenders, it has an advantage that individual lenders don’t. First, they don’t have access to other lenders’ applications. Most misrepresented data doesn’t seem fake on the surface, so without seeing that data appear on other applicant’s loans, a lender would never know it was fake.
Second, even if they did have that kind of access, they couldn’t manually review even a fraction as many loans in even 10x the amount of time. The underwriting process would get even more bogged down than it currently is. Fraudbot is brilliant at spotting discrepancies like duplicated names, addresses, social security numbers, employers, and more. It then produces alerts on any applications that exhibit patterns known to be consistent with synthetic identity fraud.
For example, Point Predictive collaborated with a lender to implement the Synthetic ID FraudBot in one of their low-risk lending channels. Even in that low-risk environment, FraudBot quickly identified that 75% of applicants using a randomized social security number in the channel were synthetic identities. FraudBot’s pattern recognition abilities further reduced the risk of one of this lender’s best-performing channels.
Imagine the results in a higher-risk channel! Fraudbot is simply the best line of defense against criminals attempting to defraud lenders.
Benefits of Synthetic Identities FraudBot
Artificial Intelligence (AI) can feel complicated or alarming in some applications but applying AI to fraud detection makes sense. Fraudsters are only getting wilier. As soon as one method of fraud is detected, they create another one. Synthetic identities work, so outwitting them without the power of AI is next to impossible. FraudBot is our best hope for finding a balance between due diligence and effectiveness in the underwriting process.
Learn more about Point Predictive’s Artificial and Natural Intelligence [AI+NI]TM approach to improving fraud detection.