Tax season is underway. With this time of year comes a spike of casework for compliance and risk teams which relate to distinct risks related to tax fraud. IRS estimates that $2.2 billion in tax fraud were committed last year, and cooperation between public and private sectors can help reduce that figure. It doesn’t have to feel daunting to your compliance and fraud teams though. There are some simple indicators the private sector could consider when detecting and investigating this type of activity. Implementations of new monitoring methods by the private sector can help our public sector partners without significantly increasing headcount for either side.

Synthetic identity fraud occurs when criminals blend real and fake customer data to create a false identity that can be used to defraud the government. This activity makes up a sizable portion of the tax fraud committed each year. To conduct synthetic identity fraud a criminal could use falsified PII, like a name, address, or email, and combine it with a valid, but stolen, social security number (SSN) either when opening an account or when updating information on that account. Once the profile is established, the fraudster can receive multiple falsified tax refunds into their account. After the refunds are deposited, the money will be debited quickly before the account can be frozen. To help identify risk earlier, private companies will monitor dark web activity for stolen credentials either in-house or by contracting a third-party service provider. If we understand what has potentially been compromised, and we know what some of the potentially suspicious actions could look like, then we can use this knowledge to detect this behavior automatically.

There are many ways we can monitor for potential tax fraud to protect our clients. We can identify compromised SSNs, look for instances of excessive tax refunds, identify changes to KYC profiles around this time of year, track irregularities in online device signatures, and track how refunds are debited from the account. But these data do not exist in a vacuum. We must look at the customer and their activity more holistically. For example, when we think about how many refunds we would consider excessive, understanding location is important. If we have information that shows someone residing only in a single state which has no state income tax, then it would seem abnormal if they received five ACH deposits from the US Treasury (which may be identified by descriptions that read “IRS TREAS 310” and include transaction details of “TAX REF”).

Updates to PII around the time of the refund(s) should also raise a red flag. But the fact that an update occurred is not sufficient to assign risk. More significant updates are cause for concern. It is much more likely for a malicious user to update multiple sensitive PII fields at once, while less significant updates are more likely to be executed by normal users. Normal users will spend their tax returns, but seeing someone immediately structuring out funds could also be cause for concern.

Building holistic models which take into account more of these pieces of information will begin to increase the proficiency with which we identify risk. If the private sector can refine its detection methods, it can provide more targeted reports with more context to the IRS.

This is key to the government recovering lost funds. Once a criminal gets their money, they withdraw it quickly. So quickly that private companies and the IRS likely don’t have time to stop them from layering and integrating the funds. During the delay between the crime and investigation, fraudsters have more time to continue perpetuating more of the same activity. The faster and more efficient the government response can be, the higher the odds of mitigation and recovery.

The IRS’s latest headcount estimate for 2021 was around 81,600 employees, and they are currently repurposing 1,200 employees to address a refund backlog. But, if private institutions start to implement some of the detection methods which target this type of risk they could more quickly investigate and disposition their cases. It would not take a large team on the private side to enhance their detection methods for this type of activity, and they would be helping their own firms improve operational efficiency.

Tax fraud season brings with it distinct risks and operational overhead for Fintechs and the IRS. But by looking at indicators in transactions, client profiles, and online behavior, the private sector can reveal some key risks at the time of the activity. In doing so, we can contribute to correctly identifying and reporting tax fraud and help contain a rapidly increasing headcount for all stakeholders.