Federal Improper Payments Oversight

• Its highlights state: 'Ghost preparers are paid tax preparers who do not comply with requirements to identify themselves on tax returns they prepare for others', and that paid preparers 'prepared more than half of all tax returns in fiscal year 2025'.
• It states: 'We identified 44 federal (21 criminal and 23 civil) ghost preparer cases involving tax fraud from 2017 through 2025.'
• On restitution in the 21 criminal cases it states courts 'ordered restitution to IRS in response to 19 cases, for a total of over $20 million', ordered taxpayer restitution 'in response to three cases, for a total of almost $16,000', and in one case ordered two co-defendants to pay nearly $62,000 to the Iowa Workforce Development.
• It states: 'IRS maintains some data on identified ghost preparers, but the systems IRS uses to track ghost preparer cases and outcomes are not connected and have limitations that impede their usefulness. As a result, IRS does not know how many ghost preparers it has identified, investigated, and addressed.'
• It states IRS's efforts 'are fragmented across six offices and IRS is missing opportunities to share information and better address risks', and that IRS 'has not developed measurable objectives or performance goals' for them.
• It describes manipulation of refundable credits in both directions: ghost preparers who fabricated dependents or circumstances to qualify a taxpayer for the Earned Income Tax Credit, and who 'reported false income to raise the taxpayers' earned income into a range that generated a higher EITC', and notes GAO's prior work on IRS's Refundable Credits Return Preparer Strategy program 'designed to address improper payments associated with refundable tax credits'.
• It states that according to IRS officials 'restitution is based on the amounts proven by the government for the specific tax returns that were charged in the indictment' and that because ghost preparers conceal their identities 'the charged ghost preparer conduct is a fraction of the overall fraud.'
• It states IRS and DOJ officials agreed 'the total effect of ghost preparer fraud is likely greater than the cases we identified because other ghost preparer cases may be charged in state court and others remain undetected by the government, and therefore never adjudicated.'
• Its highlights state GAO made three recommendations - enhancing data collection, sharing and analysis; incorporating ghost preparer efforts in a comprehensive, coordinated paid preparer strategy; and defining measurable objectives and goals - and that 'IRS agreed with GAO's recommendations and described steps it plans to take or started taking, in response to each recommendation.'

• GAO: IRS cannot say how many ghost preparers it has identified, investigated and addressed, and spreads the work across six offices with no strategy or goals →Oct 16, 2026 · audit report

A shared source does not establish conduct or a direct personal relationship.

• U.S. Government Accountability Office (GAO) →audit authority
• U.S. Internal Revenue Service (IRS) →official record holder

• It states: 'According to DOD's Office of the Chief Information Officer (OCIO), the department planned to spend $10.3 billion on the 18 major IT business programs from fiscal years (FY) 2024 through 2026.'
• It states: 'Of the 18 programs, seven programs reported via GAO's questionnaire that program staff were either unaware of or did not receive training to recognize and report signs of fraud or tampering in IT systems'.
• It states that 'while ten programs reported assessing fraud risks facing the program, five programs reported not assessing fraud risks and three programs reported not knowing if they assess fraud risks.'
• It states DOD officials indicated 'the department does not currently require training to recognize and report signs of fraud in IT systems', and that personnel instead receive mandatory, general fraud awareness training.
• Recommendation 1 asks that major IT business programs 'promote and sustain an anti-fraud tone that permeates the program's organizational culture' through fraud-risk training.

• GAO: only 10 of 18 major DOD IT business programs report assessing their fraud risks, and seven report staff without fraud-recognition training →Sep 28, 2026 · oversight finding

A shared source does not establish conduct or a direct personal relationship.

• U.S. Government Accountability Office (GAO) →audit authority

• The record reports that the $116 billion is the 5th percentile of GAO's combined model, whose 95th percentile is $253 billion, and that the $304 billion is the 95th percentile of a separate shadow-economy model.
• It reports GAO's statement that the actual amount of fraud could fall outside the estimated range.
• It reports that applied to tax year 2022 the range is approximately 2 to 6 percent of all tax owed and 17 to 43 percent of the gross tax gap.
• It reports that the IRS partially agreed with both recommendations, said the report does not sufficiently distinguish fraud from broader noncompliance, and that both recommendations are open with no deadline attached.

• GAO estimates federal tax fraud losses of $116 billion to $304 billion a year →Sep 25, 2026 · oversight finding

• The record reports that employees resolved the alerts for 20 of the 100 sampled recipients, had not started or finished the review for 63, and cleared the alert without recording all earnings for 17.
• It reports that the projected amount includes both over- and underpayments.
• It reports that about $664 million cannot be recovered under SSA's administrative finality rules absent a determination of fraud or similar fault, that about $344 million remains correctable, and that SSA agreed to all three recommendations.
• It explains that an earnings alert flags a discrepancy between the earnings a recipient reported and wage data from other sources.

• SSA OIG estimates $1 billion in SSI paid despite unresolved earnings alerts, $664 million unrecoverable →Sep 15, 2026 · audit report

• FR Doc 2026-19643, printed at 91 FR 60958-60959, published September 25, 2026; comments due October 26, 2026.
• The matching program will run for an initial term of 36 months, approximately October 19, 2026 to October 18, 2029, renewable for three years.
• It re-establishes matching begun in 2020 to address improper payments using the Do Not Pay Working System (85 FR 58062).
• CMS will provide Treasury identifying data on individual Medicare providers and suppliers from the Provider Enrollment, Chain, and Ownership System (PECOS).

• CMS re-establishes its Do Not Pay match for Medicare providers and suppliers →Sep 25, 2026 · records release

A shared source does not establish conduct or a direct personal relationship.

• U.S. Department of Health and Human Services / Centers for Medicare & Medicaid Services (HHS/CMS) →official record holder

• GAO estimated that annual federal tax fraud loss is between $116 billion and $304 billion based on data and information from 2018 through 2024.
• The range reflects approximately 2 percent to 6 percent of estimated tax owed if applied to tax year 2022.
• According to IRS, its Return Review Program prevented $88 billion in invalid, and potentially fraudulent, tax refund payments from 2018 through 2024.
• IRS has not developed an antifraud strategy or designated an antifraud entity; GAO recommended both, and IRS partially agreed with both recommendations.
• GAO reports that its 2024 recommendation for Treasury, with OMB, to expand government-wide fraud estimation was still awaiting a Treasury status update as of July 2026.

• GAO estimates federal tax fraud losses of $116 billion to $304 billion a year →Sep 25, 2026 · oversight finding

A shared source does not establish conduct or a direct personal relationship.

• U.S. Government Accountability Office (GAO) →audit authority
• U.S. Internal Revenue Service (IRS) →official record holder

• FR Doc 2026-19599, printed at 91 FR 60677, published September 24, 2026.
• The correction states the September 11, 2026 document inadvertently omitted the associated Privacy Act System of Records in the Appendix.
• The corrected Appendix states the OMB M-25-32 routine use has been added to SBA 20 — Disaster Loans Case Files (91 FR 56700) and SBA 21 — Loan System (91 FR 56703), published September 3, 2026.

• SBA corrects its Do Not Pay matching notice to name the systems of records it omitted →Sep 24, 2026 · records release

• Of 150 sampled grouped line items, 43 were unallowable because the enrollees did not meet the eligibility criteria for the target group designated in the Tennessee State plan.
• 30 of the 43 involved enrollees who had received early intervention case management services without an assessment of whether they were at serious or imminent risk of entering State custody.
• Based on its sample results, OIG estimated Tennessee claimed at least $59.3 million (Federal share) in unallowable Medicaid reimbursement.
• Tennessee did not concur with either of OIG's recommendations.

• HHS-OIG estimates Tennessee claimed at least $59.3 million in unallowable case-management reimbursement →Sep 23, 2026 · audit report

A shared source does not establish conduct or a direct personal relationship.

• U.S. Department of Health and Human Services, Office of Inspector General (HHS-OIG) →audit authority

• It states: 'For 172 of the 230 sampled enrollee-years, medical records did not support the diagnosis codes and resulted in $622,927 in overpayments.'
• It states: 'On the basis of our sample results, we estimated that United received at least $24.4 million in overpayments for 2020 and 2021.'
• It states: 'We made four recommendations, including that United refund to the Federal Government the $24.4 million in estimated overpayments.'
• The report states: 'United disagreed with some of our findings and requested that we withdraw all our recommendations.'
• The report page reads 'Issued on 09/23/2026' and 'Posted on 09/28/2026'.

• HHS-OIG estimates UnitedHealthcare of Texas received at least $24.4 million in unsupported Medicare Advantage risk payments →Sep 23, 2026 · audit report

A shared source does not establish conduct or a direct personal relationship.

• U.S. Department of Health and Human Services, Office of Inspector General (HHS-OIG) →audit authority

• OIG identified 521,393 beneficiaries entitled to student benefits for 1 or more months from January 2022 through December 2024 and randomly selected 100.
• SSA correctly paid 71 of the 100, did not correctly pay 17, and did not retain completed Forms SSA-1372 for 12.
• OIG estimates SSA improperly paid approximately $211 million to about 89,000 beneficiaries and made approximately $342 million in unsupported payments to about 63,000 student beneficiaries.
• OIG made three recommendations and SSA agreed with them.

• SSA OIG estimates $211 million improperly paid and $342 million unsupported in student benefits →Sep 23, 2026 · audit report

A shared source does not establish conduct or a direct personal relationship.

• Social Security Administration, Office of the Inspector General (SSA OIG) →audit authority