Weeding Out Fraudulent Unemployment Claims Just Became More Difficult

Scammers defrauded the state of Washington out of "hundreds of millions" of dollars by using residents’ stolen information to file claims for unemployment, the state employment office said this week.

The disclosure underscores the difficulty states face in rooting out fraudulent claims as an unprecedented number of out-of-work Americans file for unemployment because of the coronavirus outbreak.

Washington’s Employment Security Department has paid out $3.8 billion in benefits to more than 768,000 people since March, when businesses began closing down as the state tried to contain the coronavirus.

But last week the state saw a 26% jump in the number of first-time unemployment filings, and officials believe at least a portion of those are due to an increase in fraudulent claims.

“At this point, we have tens of thousands of individuals whose stolen information has been used to file fraudulent claims,” the department’s Commissioner Suzi LeVine told local reporters this week. “We believe that this translates into hundreds of millions of dollars."

An international crime ring is believed to be involved in a large scale fraud operation targeting state unemployment systems, according to a memo the Secret Service issued last week, which was reported on by the New York Times and Krebs on Security. Washington has been the primary target thus far, but authorities also uncovered evidence of fraudulent activity in North Carolina, Massachusetts, Rhode Island, Oklahoma, Wyoming and Florida and warned that every state was vulnerable to the scheme.

The Secret Service said that the crime ring behind the fraud scheme “possesses a substantial [personally identifiable information] database” and warned that  “a substantial amount of the fraudulent benefits submitted have used PII from first responders, government personnel and school employees.”

More than 38 million Americans have filed for unemployment benefits since March, when states began ordering businesses and schools to close to try and stem the coronavirus outbreak. States have struggled to keep up with the crushing demand, making it especially difficult for unemployment offices to detect and catch this type of fraud.  

“All these state systems are set up to catch or detect fraud by individual claimants. Not people working in a concerted effort to submit fraudulent claims all at once,” said Mason Wilder, a senior research specialist with the Association of Certified Fraud Examiners. 

For example, Oklahoma, like many other states, did away with a week-long waiting period before initial unemployment compensation would be paid in order to quickly get money to out-of-work residents. But that could be making it easier for fraudsters to slip through the cracks, said Kerry Master, the reemployment services and compliance division director for the Oklahoma Employment Security Commission.

“It gets assistance and relief to people faster, but it also gives the fraudulent actors a week of free time, if you will, that we would normally catch them and now don’t necessarily get that improper payment stopped immediately, but we’ll get it stopped," Kerry told KOKH-TV last month before the international fraud ring was identified.

Some of the personal information used in committing the fraudulent unemployment filings may have been acquired through prior security compromises of businesses, Wilder said. For instance, if a hacker obtained the human resources employee logs of a company, it could have the personal information and social security numbers needed to file an unemployment claim.

In that case, states may need to develop better ways to flag potentially fraudulent claims for review. Some red flags of potential fraud could include use of the same bank account for multiple unemployment claims, hundreds of claims submitted through the same IP address, or asking unemployment offices to send claims to out-of-state addresses, Wilder said.

“Those would be a red flag that should put a payment on hold until that claim is evaluated,” he said.

It could also be helpful to review claims that have been identified as fraudulent for any telltale signs that could be used to weed out others.

Some of the fraudulent activity has reportedly been traced to a Nigerian crime ring, so noting the time of day that claims are submitted (for instance, if hundreds of claims were filed at 3 a.m.) or incorrect spelling of a word in the same way could help identify other fraud, Wilder said.

Flagging a large volume of potentially fraudulent unemployment claims will also create a new set of problems.

“If they put a system in place that automates the red flag detection and they start having 5,000 cases a day that now need further human review, it’s going to be really difficult for them to handle all of that,” Wilder said of state unemployment agencies. “The biggest challenge will be from a resource perspective, the financial resources to pay for all this additional review and extra time and effort.”