Treasury's machine learning recovers $1 billion in FY2024

Treasury reported over $4 billion in fraud and improper payments prevented or recovered in FY2024, including $1 billion recovered through machine-learning check-fraud detection.

ⓘ

AI system

System

Office of Payment Integrity enhanced fraud-detection processes including machine-learning AI (Bureau of the Fiscal Service)

Model

Not assessed

Release

Not assessed

Application developer

U.S. Department of the Treasury, Bureau of the Fiscal Service, Office of Payment Integrity (no vendor is named in the sources)

Role

Not assessed

Configuration

Not assessed

Inputs and outputs

Not assessed

Autonomy

Not assessed

Workflow

Centralised disbursement data at the Bureau of the Fiscal Service (Nextgov: over 1.4 billion payments and more than $6.9 trillion a year). Machine learning applied to large payment datasets to flag potential fraud schemes, combined with risk-based screening and prioritization of high-risk transactions.

Human oversight

Not assessed

Relation to AI

Treasury's release attributes $1 billion of FY2024 recoveries to machine-learning AI that expedited identification of Treasury check fraud. It credits $2.5 billion to prioritizing high-risk transactions. The technique is not described and no vendor is named. So developer and deployer appear to be the same public body (the Bureau of the Fiscal Service's Office of Payment Integrity). This is not stated. The figures are not independently verified in the sources.

AI contribution

Treasury's release attributes $1 billion of FY2024 recoveries to machine-learning AI that expedited identification of Treasury check fraud. It credits $2.5 billion to prioritizing high-risk transactions. The technique is not described and no vendor is named. So developer and deployer appear to be the same public body (the Bureau of the Fiscal Service's Office of Payment Integrity). This is not stated. The figures are not independently verified in the sources.

Benefits

Outcome identifiers

Not assessed

Benefit

$1 billion recovered through machine-learning check-fraud detection

Category

Reduced Risk

Status

Not assessed

Beneficiary groups

Society. Organizations.

Beneficiaries

Not assessed

Reach

Not assessed

Distribution and access

Not assessed

Duration

Not assessed

Expected?

Yes

Costs and harms

Not assessed

Measurement

Studies

Not assessed

Study design

Not assessed

Measurement

Not assessed

Setting and sample

Not assessed

Measures

$1 billion recovered through machine-learning check-fraud detection

Comparison

Treasury's fraud and improper-payment prevention and recovery in FY2023, before the enhanced processes

Results

Reference: $652.7 million in FY23. Result: over $4 billion in fraud and improper payments this fiscal year (FY) (October 2023 – September 2024).

Additional value from AI

Not assessed

Limitations

Not assessed

Denominators

Not assessed

Materials for replication

Not assessed

Money

Kind of value

Treasury check fraud recovered in FY2024 through machine-learning identification

Amount

$1 billion

Currency and prices

USD, identified from the source context. Nominal source amount; price year not separately stated.

Period and unit

U.S. fiscal year 2024. Treasury check-fraud recoveries across the program.

Realization

Reported recovery.

Recipients and payers

Recipient: U.S. Treasury and affected federal programs. Payer: Recovered fraud proceeds; individual payers are not identified.

Gross or net

Gross recovered funds. Detection, recovery, and operating costs are not deducted.

Included costs

Not assessed

Comparison

The release does not estimate recoveries without machine learning.

Method

Not assessed

Return and payback

Not assessed

People and places

Participants

Developer: U.S. Department of the Treasury, Bureau of the Fiscal Service, Office of Payment Integrity (no vendor is named in the sources). Operator: U.S. Department of the Treasury, Bureau of the Fiscal Service.

Jurisdictions

Developer: United States. Operator: United States.

Ownership and control

Not assessed

Dependencies

Not assessed

Cross-border connections

Not assessed

Geopolitical implications

Not assessed

Follow-up

Conditions for replication

Access to payment and check data at scale. Operational capacity to stop or claw back flagged payments and an office that measures the results. Independent verification of the claimed figures, which the sources do not provide.

Replication attempts

1 follow-up entries. Benefit recurred: 1.

Later findings

18 September 2026. The Fiscal Service reported that Treasury's payment integrity efforts prevented, detected or recovered more than $11.7 billion in FY2025, nearly three times the FY2024 total.

Availability

Not assessed

Institutional decision

Not assessed

Grounds

Not assessed

Action and result

Not assessed

Success criterion

Not assessed

Evidence

Source numbers

AIBD-0173-source-0001–0004

Citations

Numbered references below.

Language

Not assessed

Source relationships

Developer, Independent

Original evidence

Not assessed

Access and versions

Not assessed

Supported findings

Not assessed

Evidence strength

Not assessed

Missing evidence

Not assessed

Grades

Assessed benefits

Society. Organizations.

Depth

Not assessed

Reasons

Society: The sources give only aggregate FY2024 totals for taxpayers' money recovered or prevented. No outcome is documented for an identifiable group. Organizations: Treasury reports aggregate prevention and recovery figures for the Office of Payment Integrity. No single deployment outcome is described.

Observed or intended

Not assessed

Rubric

Not assessed

Assessor and date

Not assessed

Alternative assessments

Not assessed

Fill

Independently reported

Entry details

Kind

Not assessed

Field

Public finance and payment integrity. Federal government payments and check fraud.

Purpose

Detect and prevent fraud and improper payments in federal disbursements, including Treasury check fraud.

Dates

Date basis unassessed: 2024 (precision unassessed). Updated: 21 October 2026.

Stage

Not assessed

Version

SHA-256: c2e315c046b1…

References

Other identifiers

Not assessed

Category mappings

Not assessed

Brief

Not assessed

Revisions

6 revisions. Latest: 21 October 2026.

Sources

  1. News report Independent author

    Natalie Alms, Nextgov/FCW, 18 October 2024

  2. News report Independent author

    Rob Wile, NBC News, 17 October 2024

  3. Press release Developer

    U.S. Department of the Treasury, press release, 17 October 2024

  4. Press release Developer

    Bureau of the Fiscal Service (2026). Treasury's Bureau of the Fiscal Service Recognized for Payment Integrity Efforts. Fiscal Service news, 18 September 2026.

Collections: Monetary outcomes.