Modernizing Government Payment Systems to Improve Efficiency and Effectiveness

The IBM Center for the Business of Government, in collaboration with the National Academy of Public Administration (NAPA), recently co-hosted a roundtable on Payment Integrity and Artificial Intelligence (AI).  Payment Integrity, fraud, and error reductions present important topics for Federal agencies and the new administration to address.  In April 2024, the GAO estimated that the government has cumulatively made $2.7 trillion in improper payments over the past 20 years, and $232–$521 billion is lost to fraud each year.

The roundtable brought together senior leaders from the Department of the Treasury, Department of Health and Human Services, Social Security Administration, Congress, and others who are confronting these issues.  They joined experts from industry and academia to discuss challenges and opportunities, including the use of AI to address improper payments.

Payment integrity is vital for public trust. With trillions spent annually, addressing improper payments (overpayments, underpayments, or disbursements to ineligible recipients) presents a crucial challenge.  Improper payments -- whether due to oversight, insufficient information on eligibility, or fraud -- divert resources from beneficiaries who are often economically disadvantaged, elderly, and disabled, and erode confidence in government effectiveness and fiscal stability.

Key Challenges

Improper payments have remained an issue for decades.  Roundtable participants identified key challenges that drive payment integrity issues.  These include outdated payment systems, inadequate oversight, and insufficient data analytics capabilities.  Participants also face challenges by operating in a “pay and chase model” where payments are made, a problem is identified, and recovery of payments is expensive and difficult.  Additionally, overseas cybercriminals using stolen identities to defraud multiple programs.   The illustration below highlights challenges raised in the discussion.

Issues underpinning these challenges include:

Technology and Tools to Keep Up with Bad Actors:  The government needs to keep up with nimble, increasingly sophisticated techniques to exploit vulnerabilities in payment systems.  

Identity Protection and Usability:  Identity protection and customer experience intersect in challenges like balancing security measures with user convenience, ensuring data privacy without hindering personalized service, and managing trust while preventing fraud.

Siloed Agencies and Systems:  Stovepiped programs in and across agencies can hinder real-time data exchange, foster an inability to stop problematic payments before they are made, and reduce the ability to stop bad actors from future attacks.

Outdated Laws Restricting Data Sharing:  Current privacy laws make data sharing for fraud detection very difficult, inhibiting agencies’ ability to share information about known bad fraud actors. This also slows the development of advanced algorithms for risk assessment and prevention.

Opportunities for Improvement

Participants focused on several solutions and best practices.  Addressing improper payments, especially fraud, requires federal agencies to prioritize use of AI, as well as data analytics, interagency collaboration, accountability, and workforce training. By implementing several key actions, federal agencies can safeguard payment integrity, uphold public trust, and ensure responsible use of taxpayer dollars.  The Treasury Department has introduced a new initiative around payment integrity, which can serve as one of the foundations for action.  Key concepts to drive the success of such efforts include strengthening several capabilities:

1. Collaboration: Share information and best practices across agencies, oversight bodies, private sector partners, international organizations, and other stakeholders to identify and address threats more effectively.
2. Data Analytics & Technology: Equip federal programs with tools and data to better detect and prevent fraud and improper payments. This includes leveraging AI and machine learning to detect risky transactions, reducing vulnerabilities. Sharing algorithms across organizations, rather than data, can mitigate privacy concerns. 
3. Accountability and Transparency: Conduct regular audits and utilize scorecards that tie performance metrics to payment accuracy
4. Workforce Training: Provide continuous education to staff managing payments, upholding payment integrity standards.

Next Steps

The IBM Center and NAPA are preparing a report on the specific issues and ideas raised during the roundtable.  Drawing on ideas and experiences from academia, commercial banking, and the public sector, the report will focus on key considerations and recommendations to speed progress.  We hope that the paper will serve as a resource for new leaders, agency officials, and stakeholders.