The Trump administration is moving well beyond rhetoric on tackling fraud and abuse in government benefits to institutionalizing efforts to stop it. Reports from multiple agencies spotlight the administration’s expanding deployment of big data to fight fraud and improve the accuracy of benefits. That effort includes comparing massive benefit recipient lists with databases of those who have died, are in prison, or are claiming benefits from other programs or in other states, among others. Especially given extraordinary taxpayer losses to fraud in recent years, that effort is overdue, even as significant hurdles to seamless implementation remain.
In August, the Department of Justice announced the launch of the National Fraud Detection Center (NFDC),
a prosecutor-led, multi-agency team designed to investigate the most harmful actors defrauding federal government programs, including illicit actors overseas and those operating fraud schemes across federal programs. The NFDC will bring together law enforcement agencies and analytical capabilities to generate criminal leads to drive more impactful prosecutions and enhance fraud-fighting results for the American people.
As the FBI summarized in a September review, the NFDC is designed to identify rip-offs across programs and state lines, using “an advanced AI environment to comb through vast data troves.”
Earlier this month, the Department of the Treasury, which administers refunds and tax benefits like the fraud-plagued Earned Income Tax Credit, also announced significant data improvements:
“Treasury continues to transform how the federal government protects taxpayer dollars by using better data, stronger controls, and advanced technology to stop fraud and improper payments before money goes out the door,” said Secretary Scott Bessent. “In the past year alone, Treasury built and deployed new safeguards that verified more than $3.7 trillion in federal payments and increased Do Not Pay access from 4 percent of programs to 99 percent, ensuring agencies have access to the data they need. We are moving beyond ‘pay and chase’ and making prevention the federal government’s first line of defense.”
The expansion and improved quality of the Do Not Pay program, which is administered by Treasury and “provides federal agencies and federally funded state-administered programs with data and tools to help verify identity and eligibility,” is critical. As Treasury noted, over the past year they have “streamlined access to Do Not Pay’s fraud prevention tools, added nine new data sources to the Do Not Pay program, and deployed a new government-wide payment verification process.” The result has been a “nearly four-fold increase” in records screened, from 641 million records screened in 2025 to over 2.3 billion records screened in the past fiscal year.
Recent news accounts reveal some of the fruits of that effort. On October 6, Treasury released data showing it had identified and prevented “about 13,500 payments totaling $175 million” that otherwise would have wrongly gone to dead people last year. This past week, the Department of Health and Human Services announced it was ending Affordable Care Act subsidies for about 189,000 people it identified as being deceased or also having Medicare, Medicaid, or Children’s Health Insurance Program coverage, which officials estimated would save taxpayers $1.5 billion over the next year.
In each case, that’s a fraction of the nearly $6 trillion in federal benefits, grants, and other disbursements made each year, but prevention and enforcement efforts continue growing. New verification capabilities, including over bank account ownership and Taxpayer Identification Numbers associated with federal payments, are just now coming online and are expected to expand agency capacity to prevent future improper payments.
Still, basic challenges remain. In June, the nonpartisan Government Accountability Office (GAO) reported that the success of data matching ultimately depends on “data interoperability—including standardized data elements and increased interoperability of data elements, such as names and addresses across data sources and agencies.” GAO noted that issues in multiple datasets “undermine data reliability and interoperability for agencies seeking to make eligibility determinations.” In the end, GAO called on Congress to act “by assigning a single agency a lead role in establishing and implementing data interoperability requirements for recipient eligibility data sources.” Based on its already central role in paying tax benefits and administering the Do Not Pay system, GAO suggested Treasury would be a logical choice.
Congress and the administration should ensure such interoperability applies to permanent programs and temporary programs created in response to recessions, which were widely abused during the pandemic. For example, the Department of Labor admitted in August 2023 that the federal Pandemic Unemployment Assistance program experienced a staggering 36 percent improper payment rate, which understated the true degree of taxpayer losses.
