How automation helps agencies stop emergency relief fraud

Local, state and federal agencies are uniquely positioned to provide vital relief funds to communities in crisis. These organizations are under significant pressure to quickly distribute large sums of money to meet urgent needs. Their teams are often small and spread thin, making aid programs susceptible to fraud. 

When funds are misappropriated, people don’t get the help they need, and public trust in government programs is undermined. Fortunately, technology is evolving to help public-sector funders strengthen how they manage relief funding programs, significantly reduce fraud and move more efficiently. But technology is only effective with the right people at the helm. 

A tighter, cleaner review

Automation is a critical tool in helping public funders ensure emergency relief funds reach the intended recipients quickly. Incorporating automation into the review process allows funders to streamline and strengthen their reviews without burdening applicants. 

Currently, there are two main approaches to anti-fraud automation: knowledge-based authentication and identity verification. Knowledge-based authentication verifies the identity of grant and relief-fund applicants at specific points in the application process. It consists of a series of questions unique to the private and personal history of applicants—their mother’s maiden name, the street address of their childhood home—like a password retrieval process. Dynamic knowledge-based authentication goes one step further, generating questions based on public and private data sources, such as marketing data, credit reports or transaction histories specific to the applicant. By adding a few simple questions, funders can verify an applicant’s identity. 

Identity-verification capabilities also help public-sector funders confirm that an applicant uses a legitimate government-issued photo ID. Smartphones provide a big assist here, as applicants are prompted to submit a photo of their government-issued ID followed by a live selfie to verify they match. 

Services teams can employ automation in other ways to improve the application review process, including validating revenue and tax documents. For instance, applicants can upload their W2s to verify their annual earnings through a process that connects with the state’s department of revenue database. The team can scan the applicant’s W2 via assisted review and validate what the applicant provides against the state’s records. 

Automated technologies can also help services teams establish IP fencing to validate that an applicant’s IP location falls within an aid program’s geographical boundaries, such as for a state grant program. Tech-assisted review can efficiently enforce deduplication to ensure that applicants only submit one application. Finally, funding organizations can digitize their scoring rubric. To do so, the review team trains and tests a scoring model to mirror how the team scores applicants, ensuring it accurately reflects the human-generated scores. 

Streamlining the review processes enables government teams to deliver large-scale relief swiftly and prevents them from trying to manually root out fraud—an impossible task. 

The importance of the human element

Though technology is a powerful tool, it must be paired with strong management and effective safeguards. Technology is not a replacement for the human teams who run relief programs. 

For example, in Ohio, a teleservices representative with an HR consulting firm allegedly removed fraud flags on applications for pandemic unemployment assistance and received kickbacks. Though she was fired from her job, investigators found she was still able to unflag 176 fraud issues for at least a month after being terminated. 

In this case, an automated system was used, but without the proper human management and safeguards to ensure a former employee could no longer exploit the system. This oversight resulted in at least $166 million in fraudulent applications reportedly being paid out in Ohio between July 2020 and June 2021 (and this doesn’t consider 2022 figures, or the $1.2 billion in benefit payments flagged as potentially fraudulent that are still being adjudicated).

Putting automation into action

Alternatively, the recent launch of a U.S. state payment program for frontline workers shows the power that automated technology brings to public-sector funders and community members. 

In May 2022, a state government agency prepared to launch the payment program for frontline workers who supported their communities during the COVID-19 pandemic. The state legislature gave the agency a 45-day window to launch. To move quickly, the state agency turned to automated software-as-a-service (SaaS) technology to mobilize the program. In doing so, they launched the program in under one month and reviewed more than 1 million applications received within the first 45 days. 

The technology delivered a simple, streamlined and equitable user experience and ensured that only legitimate applicants moved forward in the review process. With specific criteria for applicants, the team could instantly weed out ineligible or fraudulent submissions. Applicants deemed ineligible were given 15 days to appeal for a final pass-fail determination from the state to ensure that no one was denied aid because of an application mistake. 

Using tech-assisted review, the program launch timeline was approximately halved from the eight weeks initially expected, delivering essential support to frontline workers quickly. After the program’s success, the state agency will continue to use SaaS technology to distribute $500 million of aid to over 800,000 eligible workers. It will also use the solution to distribute 1099 tax forms to all eligible applicants. 

Ensuring aid lands in the right hands

Large-scale relief fund management brings many challenges. Government teams must be ready to review countless applications, confirm their validity and make sure large sums of money reach intended recipients swiftly. They must have the capacity to launch programs in a matter of weeks and protect public funds from fraud while ensuring equity and giving applicants a seamless experience. 

Unfortunately, as long as government aid programs exist, fraud will continue to be a threat. Empowering a human services team to leverage automated technology can enable agencies to launch, manage and measure social impact programs more effectively and with reduced risk. 

Sam Caplan is the vice president of social impact at Submittable.