May 24, 2018
Synthetic identity fraud is a form of identity fraud that is gaining popularity as fraudsters look to establish credit, open deposit accounts, as well as obtain driver’s licenses and passports. Creating new identities includes a combination of real and fabricated information, or sometimes entirely fictitious information.
Because fraudsters are always looking for new strategies and loopholes to steal from organizations, synthetic identity fraud is growing. By how much? Javelin Strategy & Research reported that synthetic identity fraud registered approximately $16.8 billion in 2017. This type of fraud is perpetrated on weaknesses within the banking, credit rating agencies and credit issuers reporting structure.
Running a synthetic identity fraud scam requires patience and time. Armed with a new credit account, fraudsters will legitimately use the credit account and make payments to establish good history. The fraudster will leverage the positive credit history to obtain more credit cards, retail store credit accounts and potentially car loans. This results in a compounding amount of stolen dollars. The victims become both the card issuers and merchants, as well as the individual that had some portion of their personal data stolen to help establish the false identity.
To address this growing threat, the financial market is responding in a variety of ways. Some community banks are demanding that individuals physically come into the regional bank to engage with them on a one-to-on basis. This in person experience can often stop the bank from becoming a victim. At larger financial institutions, the idea of building a rapport with an account manager is more difficult. They are often armed with more sophisticated fraud and risk solutions, leveraging artificial intelligence, and more specifically the machine learning capabilities of their enterprise fraud platform, to help stop fraud. Similar to stopping card not present fraudulent transactions, machine learning technology can help fraud analysts assimilate data, associate geographical location information and recognize patterns or anomalies as it relates to the authenticity of the credit applicant.
Enterprise-level fraud prevention solutions are armed with an assortment of patented and proprietary technologies to help both the card issuers and merchants. Machine learning solutions will help analyze transactional and contextual data to assess risk. These are just a few technologies and processes that feed data into an enterprise fraud solution to help identify and stop fraud:
- Multi-Layer Device FingerprintingTM. The process of analyzing customer behavior associated with a device – desktop, tablet, smartphone, game console, etc. – to capture a digital identity or fingerprint.
- Proxy Piercing. Determining if the transaction request is being relayed through a proxy to find the true geolocation and type of network being used.
- Mobile Device Analysis. Technologies that can help identify the specific mobile device type that is being used to complete a purchase as well as attributes of that specific device (operating system and version, browser type).
- Order linking. Drawing associations across a variety of different data points that are outside a merchant’s network to uncover fraud trends that may not have been seen when looking at the individual transactions.
- Dynamic scoring. Continuously monitoring and updating a score based on new patterns that exhibit signs of fraud.
- Historical data. Feeding historical data into machine learning algorithms enables it to learn from a merchant’s unique risk threshold to identify patterns and trends that influence business decisions.
- Machine Learning. Machine learning, an application of broader artificial intelligence (AI) techniques, trains computer systems to learn and improve from experience without being explicitly programmed.When applied to stopping fraud, machine learning provides the analytic horsepower to analyze the massive amount of data and make recommendations aligned with the specific merchant’s business objectives.
As each card issuer looks to address synthetic identity fraud in their own way, it will require the industry as a whole to work together and understand the breakdowns to address thisgrowing problem of identity fraud.
Twenty-five years ago, major banks established Early Warning Services (EWS) to monitor, collect and report on the consumer banking habits. This system is still in use today and allows banks to easily exchange information between organizations to identify and thwart fraud. With both banks and credit unions as EWS subscribers, customers are screened for prior history of fraud, forgery, account abuse, counterfeiting, identify verification, account owner authentication, etc. EWS is an example of how the industry came together 25 years ago to stop fraud.
As synthetic identity fraud reaches a tipping point, leveraging a solution that employs machine learning, customized policy rules, and domain expertise with longevity in the industry, will be vital to staying ahead of the fraudsters.
Read why it takes more than a class of algorithms to prevent fraud in the eBook “The Truth About Machine Learning in Fraud Prevention“.