evidoria

← Back to browse

Good practice Imported

US Treasury Office of Payment Integrity — Machine learning to prevent and recover federal payment fraud

United States of America · Washington · See the United States of America profile · See the Washington profile

Evidence: Observational / pre–post Top 12% 80/100 · Ask Evidence Copilot about this practice

The US Treasury's Office of Payment Integrity uses ML to screen federal disbursements. In FY2024, ML-driven processes prevented or recovered $4.059 billion — a 6× increase from $652.7 million in FY2023 — by flagging high-risk transactions and accelerating Treasury check-fraud det

4.059 USD billion
Fraud/improper payments prevented or recovered (FY2024)
652.7 USD million
Fraud/improper payments prevented or recovered (FY2023)
6x
Year-on-year increase (FY2023 to FY2024)
US Treasury Office of Payment Integrity — Machine learning to prevent and recover federal payment fraud

Details

Maturity
Scaling
Promoter
Office of Payment Integrity, Bureau of the Fiscal Service, US Department of the Treasury
Period
2023–present
Keywords
government, financial integrity, fraud detection, machine learning, federal payments

Context

The Office of Payment Integrity (OPI) within the Bureau of the Fiscal Service, US Department of the Treasury, applies machine learning models to screen federal disbursements in real time and flag high-risk transactions before settlement. OPI's tools are available to other federal agencies and federally-funded state programmes.

Objectives

To prevent and recover federal payment fraud and improper payments using ML-based screening and accelerated Treasury check-fraud detection.

Activities

ML models screen disbursements in real time; results were officially announced by the US Treasury in October 2024 with an itemised breakdown by ML method.

Results

In fiscal year 2024, ML-based fraud prevention and accelerated check-fraud detection resulted in $4.059 billion prevented or recovered — a sixfold increase over FY2023's $652.7 million.

Conclusions

The programme demonstrates how ML applied to government payment data at scale can produce measurable, independently reported fiscal outcomes with a clear year-on-year performance trajectory.

Implementation

Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful