What GAO Found State and local government recipients of the Coronavirus State and Local Fiscal Recovery Funds (SLFRF) program had obligated nearly all of their SLFRF awards, as of March 31, 2025—the most recent data available at the time of this review. The data include information received by Treasury through September 30, 2025. Specifically, states reported obligating all but $10.4 million of the $195.8 billion they received, and localities reported obligating all but $101 million of the $127.8 billion in awards they received. State and local government recipients generally had until December 31, 2024, to obligate the SLFRF awards they received. Funds that have not been obligated by that date generally must be returned to Treasury. Between March and October 2025, Treasury sent instructions to the state and local recipients that reported unobligated funds, requesting the return of those funds. As of November 2025, states and localities had returned $13.7 million of the total reported $111.4 million in unobligated funds. Further, states reported spending 80 percent ($156.3 billion) of their SLFRF awards, and localities reported spending 84 percent ($107.2 billion) of their awards as of March 31, 2025. Reported Spending of Coronavirus State and Local Fiscal Recovery Funds (SLFRF) by States and the District of Columbia, as of March 31, 2025 Both states and localities reported spending most of their SLFRF awards—53 percent ($82.6 billion) and 67 percent ($71.9 billion), respectively—to replace revenue lost due to the pandemic. States and localities generally have until December 31, 2026, to spend their awards. Why GAO Did This Study Overall, SLFRF allocated $350 billion to tribal governments, states, the District of Columbia, local governments, and U.S. territories to help cover a broad range of costs stemming from the health and economic effects of the COVID-19 pandemic. SLFRF recipients must regularly submit reports to Treasury on their use of SLFRF awards and the projects undertaken with them. The CARES Act includes a provision for GAO to monitor the use of federal funds to respond to the COVID-19 pandemic. This report examines the SLFRF funding states and localities are required to report to Treasury. This report updates our October 2023, April 2024, and September 2024 reports on states’ and localities’ spending and uses of SLFRF. To conduct this work, GAO analyzed Treasury SLFRF project and expenditure data and interviewed Treasury officials. GAO also reviewed relevant federal laws and regulations governing the SLFRF program and Treasury SLFRF program guidance, policies, and procedures. For more information, contact Jeff Arkin at arkinj@gao.gov.  

What GAO Found The Army, Navy, and Air Force have issued guidance to help facilitate command efforts to locate a service member who is deemed absent from their assigned duty location. However, the Marine Corps has not developed such guidance, as GAO recommended in 2022. GAO analyzed Army, Navy, and Air Force guidance and identified the following gaps that could hinder efforts to locate absent service members and mitigate related risks. Response time frames. Service guidance outlines response time frames with varying levels of specificity, resulting in different interpretations among officials regarding how quickly certain actions should be initiated. For example, Army guidance includes detailed time frames for actions such as alerting law enforcement, whereas Navy and Air Force guidance does not. Mental health. Army, Navy, and Air Force officials GAO interviewed commonly observed a link between service member absences and mental health and said that locating service members often intersects with efforts to prevent self-harm. However, guidance inconsistently addresses the interconnected nature of mental health issues and service member absences, and how such considerations should inform the command’s response. Safety. Army, Navy, and Air Force officials identified potential safety issues that may arise while searching for an absent service member, especially if the service member is experiencing a mental health crisis or has access to a firearm. However, guidance does not address these safety issues, potentially subjecting the absent service member or those trying to locate them to unnecessary risk. By addressing these gaps in guidance, the services can better position themselves to help prevent harm and save lives. Some services’ guidance for commanders and the military criminal investigative organizations (MCIO) lacks clarity on whether and when to classify an absence as voluntary or involuntary, which can significantly affect the urgency and comprehensiveness of search efforts. For example, Army guidance for commanders requires them to presume the service member is potentially in danger and to presume the absence is most likely involuntary after 48 hours unless available information indicates the absence should be considered voluntary. However, Department of Defense (DOD)-wide guidance does not have a similar provision, nor do the other military services’ guidance. In another example, Air Force MCIO guidance requires investigators to treat all absences as involuntary in the first instance, while guidance for the Army and Navy MCIOs does not. By revising guidance, commands and MCIOs will have a more consistent approach to absences and further their goal of quickly and safely locating absent service members. Why GAO Did This Study When a service member is absent from their unit, it may not be immediately clear if the absence is voluntary—that is, deliberate on the part of the service member—or involuntary, meaning the service member may be in danger. A timely and well-coordinated response to a service member’s absence is critical to establishing the facts and helping to ensure their safe return, if possible. House Report 118-125 includes a provision for GAO to review policies and procedures related to missing and absent service members. This report builds on GAO’s 2022 report on this topic and examines the extent to which DOD and the military services have clarified guidance for responding to incidents of absent service members, among other issues. GAO reviewed DOD guidance on responding to service member absences. GAO also visited a nongeneralizable sample of eight military installations, two per military service; reviewed the processes for responding to service member absences; and interviewed officials responsible for responding to such absences.

Why This Matters The Department of Homeland Security’s (DHS) U.S. Customs and Border Protection (CBP) has primary responsibility for securing the 4,000-mile border between the United States and Canada. Border Patrol, within CBP, reported that apprehensions in this region more than tripled from fiscal year 2019 through fiscal year 2024. As we reported in 2024, the agency has not met agent staffing targets in recent years. GAO Key Takeaways Border Patrol’s efforts to secure our nation’s borders include apprehending people suspected of illicit activity such as entry without inspection and drug smuggling. Apprehensions and drug seizures. From fiscal year 2023 to 2024, the number of people Border Patrol apprehended along the northern border increased sharply (see fig.). From fiscal year 2019 through fiscal year 2024, the number of Border Patrol’s drug seizures in this region varied. Technology. CBP uses aircraft, vessels, and surveillance technology—such as cameras, radar sites, and sensors—as part of its efforts to secure the northern border. From fiscal year 2019 through fiscal year 2024, CBP’s deployment of this technology increased. Staffing. In this same 5-year period, the number of agents staffed along the northern border decreased, but CBP has initiatives underway to address this issue. In addition, there was a decrease in the staffing rate for Law Enforcement Information Systems Specialists who monitor surveillance technology. The staffing rate for this key position along the northern border has been below its target, and the agency does not have a plan with strategies to address the staffing gap. Developing such a plan could help Border Patrol better carry out its responsibility to secure the northern border. Border Patrol Apprehensions Along the Northern Border from Fiscal Year 2019 Through the First Half of Fiscal Year 2025 How GAO Did This Study We analyzed Border Patrol data on apprehensions and drug seizures, as well as CBP data on staffing and resources since fiscal year 2019. We visited six CBP units along the northern border, selected based on apprehension levels, among other factors. We also interviewed CBP officials from the other units along the northern border.

What GAO Found Under a 3-year pilot program, the Department of the Treasury’s Do Not Pay system has temporary access to the Social Security Administration’s (SSA) full Death Master File. Prior to the pilot, the system had access to a less-comprehensive version of the file which excluded state death records. In January 2026, Congress passed a bill that, if enacted, would make access permanent. In April 2025, Treasury reported that the first year of this pilot (calendar year 2024) resulted in identification and prevention or recovery of $113.5 million in improper payments. This amount, offset by $4.6 million in costs, represented a return on investment of about 23 times Treasury’s pilot costs. Treasury projects that the pilot will result in over $337 million in net benefits over 3 years. Death data collected by states are their property, and statute requires SSA to pay states for use of these data. In September 2023, SSA concluded negotiations on new contracts to obtain states’ death data. Under these contracts, SSA paid states $23.8 million in 2024, a significant increase from 2023. SSA’s Actual and Estimated Future Costs for State Death Records, 2023–2028 The Social Security Act, as amended, specifies that SSA shall reimburse states for specific costs associated with death data, including (1) a fee for use of the data and (2) the full documented cost of transmitting the data to SSA. However, GAO found that SSA did not obtain the required state cost information and therefore did not consider it during negotiations. Instead, SSA and the states agreed on a fee structure based on the timeliness of submission of death records. Agencies receiving state death data must pay SSA a proportional share of its costs in obtaining the data from states. Of the $25.9 million in estimated state death record costs for 2025, SSA’s proportional share is projected to decline, from 42 percent in 2024 to 23 percent in 2025, due to a methodology change. Specifically, SSA calculated its 2025 share based on the percentage of SSA’s total federal outlays rather than on costs of obtaining data. SSA estimated that its outlays were about 23 percent of the federal total, so it decided that would be its cost share for 2025. For the remaining 77 percent, SSA distributed 36 percent to Treasury and 41 percent to other agencies without regard to cost (each of the other agencies is to contribute the same amount, approximately $1.533 million). Why GAO Did This Study Improper payments remain a long-standing and significant problem in the federal government. GAO previously reported that one strategy to help prevent improper payments is up-front verification of eligibility through data sharing and matching. Agencies can verify the eligibility of applicants through Treasury’s Do Not Pay system—a centralized data-matching service for agencies to use in preventing and detecting improper payments. This service currently includes the full Death Master File, SSA’s compilation of deceased Social Security number holders, to help agencies prevent improper payments. This report (1) describes the first-year results of the 3-year pilot to include SSA’s full Death Master File in Treasury’s Do Not Pay system, (2) evaluates the extent to which SSA’s payments for state death data are based on states’ documented costs, and (3) evaluates the extent to which SSA is charging agencies a proportional share of its costs. GAO analyzed SSA and Treasury documentation and interviewed federal and state agency officials.

What GAO Found The Coast Guard has over 40,000 active-duty military personnel who often reside in communities near their Coast Guard duty stations. Coast Guard units are often situated along major waterways and coastlines in remote or high vacation rental areas. Coast Guard members and their families may experience incidents of hostility, harassment, and discrimination from members of their community. Coast Guard members reported 112 incidents of community hostility, harassment, or discrimination directed towards them or their family from fiscal years 1998 through 2024. Most of these incidents, which the Coast Guard calls social climate incidents, were perceived to be motivated by race or ethnicity. More than half of these incidents were reported from fiscal year 2019 through fiscal year 2024. For example, in one incident a Coast Guard member reported that a restaurant refused to provide them service because of their race. In another incident, a service member reported that they were subjected to racial slurs by a community member in a grocery store. Reported Perceived Motivation in Coast Guard Social Climate Incidents From Fiscal Years 1998 through 2024 GAO found that Coast Guard commands have generally followed the social climate incident policies outlined in the Coast Guard Civil Rights Manual. However, district commands reported different interpretations of the definition, specifically as to whether social climate incidents are limited to legally protected classes. As a result, the Coast Guard may not have implemented its definition of a social climate incident consistently. By clarifying its definition of these incidents, the Coast Guard could better ensure commands apply policies consistently. Furthermore, the Civil Rights Directorate (CRD) was unable to locate any documentation for six of the 30 reported social climate incidents from fiscal years 2020 through 2024 that GAO selected to review. Developing and implementing a standardized process for the collection and retention of documents would help the CRD better oversee and manage the response to social climate incidents. The CRD provides Coast Guard members access to an internal website that tracks all reported social climate incidents and displays incident trends, including the location and type of incident. However, seven of the nine Coast Guard district commands GAO interviewed were not aware of this tool. By ensuring that all commands are consistently made aware of available tools, the Coast Guard could better prepare them to prevent and respond to incidents related to the social climate in the communities they serve. Why GAO Did This Study The Coast Guard mandates that all personnel be treated fairly and with respect to successfully carry out its missions, including within the communities in which they live. According to the Coast Guard, social climate incidents can have a negative impact on service members and their families. Under the Coast Guard’s Civil Rights Manual, the CRD is the headquarters office responsible for overseeing the management of reported social climate incidents. According to the Manual, one of the CRD’s strategic goals is to conduct activities and develop tools that assist and support commands in proactively preventing unlawful discrimination. GAO was asked to examine policies and procedures that the Coast Guard has in place to track, monitor, and address social climate incidents. This report provides information on available data on social climate incidents, the extent to which the Coast Guard follows social climate incident policies and procedures, and how the Coast Guard responds to social climate incidents. GAO reviewed relevant laws and Coast Guard policies and procedures, analyzed Coast Guard data, and interviewed agency officials. Specifically, GAO interviewed CRD officials, including Civil Rights Service Providers. GAO also interviewed commands from all nine Coast Guard districts as well as commands from four of 36 Coast Guard sectors and selected units within those sectors. GAO reviewed available Coast Guard data on social climate incidents reported from fiscal years 1998 through 2024 and analyzed the underlying documentation for a non-generalizable sample of 25 social climate incidents reported from fiscal years 2020 through 2024.

What GAO Found DOE established the Office of Clean Energy Demonstrations (OCED) in December 2021 to manage the historic amount of funding appropriated to DOE for demonstration projects. Key responsibilities include managing projects and assessing lessons learned to improve project oversight. OCED projects vary in scale and cover diverse clean energy technologies, including hydrogen and advanced nuclear energy. Generally, at least 50 percent of the project costs are borne by the project awardees. As of November 2025, OCED had committed over $18 billion to about 100 projects, although 35 of these projects have been identified for termination. To manage projects, OCED developed a framework that included a phased project approach and independent assessments at key oversight points to reduce organizational biases. Previously, OCED estimated it would need 351 staff to be fully staffed for the expected number of demonstration projects. There have been significant changes at OCED in 2025 affecting its capacity, including a significant decrease in workforce and contract support. OCED lost 85 percent of its staff (from 285 staff to about 40) between January and June 2025. This included all independent assessment staff. Further, in February 2025, OCED issued a stop work order on almost all contracts supporting the office. Office of Clean Energy Demonstrations (OCED) Workforce Changes from January 2025 to June 2025 and OCED’s Committed Funds and Projects as of November 2025 Currently, it is unclear how DOE plans to meet its statutory requirements to manage projects given OCED’s limited capacity. The office is unlikely to be able to conduct in-depth project reviews at key oversight decision points and an OCED official said the office will likely move some aspects of OCED’s oversight to other DOE offices. For example, the Office of Project Management agreed to conduct independent assessments, but that office lost about 60 percent of its staff in 2025. Without a plan to meet the statutory requirements of managing projects and assessing lessons learned, DOE faces increased risks of not having the capacity to manage and oversee billions of dollars of federal funding for demonstration projects. Why GAO Did This Study The Infrastructure Investment and Jobs Act (IIJA) of 2021 requires the Department of Energy (DOE) to establish a program for clean energy demonstration projects. Roughly $27 billion was appropriated to DOE for large-scale clean energy demonstration projects. Congress included a provision in the IIJA for GAO to examine the oversight of demonstration projects and recommend changes for the purpose of better carrying out the program. This report examines (1) recent changes at OCED and (2) DOE’s capacity to successfully manage projects given these changes.  GAO reviewed DOE and OCED documents, including policies, guidance, and award documentation. GAO also interviewed DOE officials and stakeholders, including surveying 45 recipients of OCED awards.

What GAO Found The Federal Bureau of Prisons (BOP) does not know how many individuals are currently in prison that could have already transferred to home confinement or a residential reentry center (RRC), also known as a halfway house. BOP officials said they do not know because the dates individuals are eligible to transfer are not readily available. GAO found that some individuals have remained in federal prisons despite being eligible to relocate to home confinement or an RRC. For instance, GAO found that BOP did not apply all the earned time toward placement in RRCs and home confinement for 21,190 of 29,934 individuals reviewed, for reasons such as insufficient RRC capacity and court orders. However, the full scale of this issue is unknown due to the lack of readily available data on eligibility dates. Until BOP maintains and monitors such data, it cannot ensure individuals transfer on time and take corrective action when timely transfers do not occur. As a result, BOP cannot ensure individuals receive the services and have the opportunities available at an RRC or home confinement, such as finding employment and long-term housing and reconnecting with the community. BOP has reported that such services can also help reduce recidivism. Limited capacity in BOP contracted RRCs and home confinement spaces was a reason that individuals did not transfer on time, according to BOP officials. However, BOP does not know the full extent of this shortage because it has not comprehensively assessed its capacity and related budgetary needs. Without these assessments, BOP cannot ensure it has enough space for incarcerated individuals to transfer on time. BOP could also miss opportunities to increase revenues and decrease costs to the federal government. For instance, BOP said that individuals who have resided in an RRC are less likely to return to prison. GAO also found that BOP made roughly 65,000 late payments to contractors, including RRCs, from fiscal year 2022 through March 2025. As a result, the agency paid $12.5 million in interest penalties as part of $2.8 billion in payments to contractors. In addition, GAO found that BOP paid RRCs late about 70 percent of the time, from fiscal years 2023 through 2024. RRC staff said they face hardships due to the late payments—needing private loans to pay staff. One RRC representative said late payments have made some RRCs reluctant to bid for new BOP contracts, which can further complicate BOP’s plans to expand capacity. By implementing a corrective action plan to address its late payments, BOP could save federal funds and better position itself to expand RRC capacity. BOP’s Late Payments to RRCs and Other Contractors, October 2021–March 2025 Why GAO Did This Study BOP contracts with roughly 150 RRCs across the U.S. to help incarcerated individuals reenter their communities upon completion of their sentences. RRCs facilitate reentry services (e.g., employment services, drug treatment, and classroom education) to individuals who reside in RRCs or who are on home confinement. RRCs can help individuals rebuild ties to their community and reduce the likelihood that they will commit future crimes. GAO was asked to review BOP’s use of RRCs. This report examines, among other things, how many individuals in BOP custody are eligible to transfer to RRCs and home confinement; the extent BOP knows its RRC capacity needs across the U.S.; and the extent BOP has paid RRCs and other contractors on time. GAO reviewed relevant federal laws, BOP policies and documents, and BOP data on RRCs, including payments to contractors. In addition, GAO selected seven RRCs and three BOP field offices and interviewed residents and staff. GAO selected locations based on criteria such as geographic dispersion and the size of RRCs within an area. GAO also interviewed BOP officials responsible for residential reentry management and oversight.

What GAO Found Each year, the Department of Homeland Security’s (DHS) U.S. Citizenship and Immigration Services (USCIS) processes millions of applications and petitions for persons seeking to visit or reside in the U.S. or become citizens. USCIS's Fraud Detection and National Security Directorate (FDNS) leads the agency’s efforts to combat fraud. In September 2022, GAO reported that USCIS could better ensure its antifraud efforts are effective and efficient by taking a strategic and risk-based approach that aligns with leading practices in three areas of GAO’s Framework for Managing Fraud Risks in Federal Programs (see figure). Specifically, GAO found that USCIS had conducted fraud risk assessments for a small number of specific immigration benefits, but did not have a process for regularly conducting those assessments. In addition, GAO reported that FDNS had not developed an antifraud strategy to guide the design and implementation of antifraud activities, as well as the allocation of resources to respond to its highest-risk areas. GAO also found that FDNS had not evaluated its antifraud activities for efficiency and effectiveness. Taking action to implement these leading practices will help USCIS ensure that it is effectively preventing, detecting, and responding to potential fraud. USCIS Should Take Action in Key Areas to Manage Fraud Risks In December 2025, GAO reported that from May 2022 through September 2024, about 774,000 noncitizens were granted parole—temporary permission to stay in the U.S.—across three humanitarian parole processes for noncitizens with U.S.-based supporters. These processes included Cubans, Haitians, Nicaraguans, and Venezuelans; Uniting for Ukraine; and family reunification parole. USCIS was responsible for reviewing supporter applications for evidence they had sufficient financial means to support prospective parole beneficiaries and met other requirements. In early 2024, FDNS officials analyzed 2.6 million supporter applications and found that fraud risks were widespread in Uniting for Ukraine and Cubans, Haitians, Nicaraguans, and Venezuelans parole processes. For example, fraud indicators included supporter information belonging to deceased individuals and thousands of applications with at least one piece of fictitious supporter information. USCIS attributed the fraud risks to insufficient internal controls in its supporter vetting process—for example, not having automated processes to prevent or detect possible fraudulent activity. DHS has since suspended or terminated the processes. However, GAO found that USCIS has not developed an internal control plan for new or changed programs in the future. Such a plan could include basic antifraud controls and mechanisms to help proactively identify and mitigate fraud risks. Why GAO Did This Study Granting immigration benefits to individuals with fraudulent claims can jeopardize the integrity of the immigration system by enabling individuals to remain in the U.S. and potentially apply for certain federal benefits or pursue a path to citizenship. In 2022 and 2023, DHS introduced new processes for humanitarian parole in response to increases in noncitizens arriving at the southwest border. The processes allowed eligible noncitizens from certain countries to travel to the U.S. to seek a grant of parole. To be eligible, noncitizens had to have a U.S.-based supporter apply to financially support them. As of December 2025, DHS had suspended or terminated the processes. This statement discusses USCIS efforts to manage fraud risks (1) across immigration benefits it adjudicates and (2) in humanitarian parole processes for noncitizens with U.S.-based supporters. This statement is based primarily on our September 2022 and December 2025 reports on these topics.

What GAO Found The Marine Corps and Air Force had higher rates of the most serious accidents in the last 2 fiscal years (2023 and 2024) with its Osprey aircraft compared with the average serious accident rate for the previous 8 years, based on available Department of Defense (DOD) data. In fiscal years 2023 and 2024, 18 serious, non-combat accidents occured involving death; permanent disability; extensive hospitalization; property damage of $600,000 or more; or a destroyed aircraft. Rates of serious accidents were between 36 percent and 88 percent higher than each service's average rate for the prior 8 fiscal years. Percent Difference of Serious Osprey Accident Rates in Fiscal Years 2023 and 2024 Compared to Service Average for Fiscal Years 2015–2022 Note: The accident rate equals the number of accidents per year divided by the number of flight hours per year and then multiplied by 100,000. Serious accidents refer to combined Class A and B accidents which are those accidents that involved death; permanent disability; extensive hospitalization; property damages of $600,000 or more; or a destroyed aircraft. The Navy had not experienced a Class A or Class B accident with its Osprey variant since it began operational use in fiscal year 2021 through fiscal year 2024. Rates of serious accidents with the Osprey aircraft generally exceeded those of the Departments of the Navy and Air Force fixed wing and rotary wing aircraft fleets for fiscal year 2015 through fiscal year 2024. Serious Accident Rate Comparisons for Marine Corps and Air Force Osprey with Departments of the Navy and Air Force Fixed Wing and Rotary Wing Fleets, Fiscal Years 2015–2024 Note: Serious accidents refer to combined Class A and B accidents which are those accidents that involved death; permanent disability; extensive hospitalization; property damages of $600,000 or more; or a destroyed aircraft. Most reported causes for serious accidents related to materiel failure of airframe or engine components and human error during aircraft operations or maintenance, according to GAO's analysis of safety data. Osprey program stakeholders have not fully identified, analyzed, or responded with procedural or materiel mitigations to all safety risks. For example, program stakeholders, which include the Osprey Joint Program Office and military services that operate the aircraft, had closed 45 risk assessments at the time of our review, but had not fully responded to 34 known system-related risks related to the potential failure of airframe and engine components. Stakeholders also had not identified actions to respond to non-system safety risks associated with aircraft maintenance and operations, such as mismatches in maintenance skill levels and limited training time to build aircrew experience. Without refining the joint program's process for identifying, analyzing, and responding to Osprey safety risks to incorporate and prioritize system and non-system safety risks, program stakeholders cannot adequately mitigate risks that can contribute to death, injury, or loss of mission capability and resources. Program stakeholders described factors that affected their ability to fully resolve Osprey safety risks, including challenges with how program stakeholders developed priorities and plans for addressing safety issues. These stakeholders identified costly, long-term engineering solutions and safety improvements that were required across a joint program with varied fleet sizes, which created longer risk mitigation timelines. GAO found that the median age for 28 unresolved serious and medium system risks was about 9 years, and over half (17 of 28) had been unresolved for between 6 and 14 years. Summary and Median Age of Unresolved Osprey System Safety Risk Assessments, by Assessment Type as of June 2025 Note: The Department of Defense (DOD) designates risk assessments as serious and medium based on their assessment of the severity (e.g., catatsrophic) and frequency (e.g., remote). The figure does not include six additional risk assessments for general military aviation risks (e.g., bird strikes) that are not specific to the Osprey and have been accepted for the life of the program. New initiatives established in 2024 are intended to address several of these factors, but these efforts did not comprehensively address non-system safety risks or include information for each service's Osprey variant. Without determining an oversight structure with clearly defined roles and responsibilities for resolving known safety risks or conducting periodic reviews of efforts to resolve them, DOD cannot have reasonable assurance that program stakeholders will fully resolve the interrelated system and non-system safety risks affecting the Osprey. GAO found that the Osprey program stakeholders have not routinely shared information in three areas to promote the safe operation of the aircraft.  Hazard and accident reporting. Program stakeholders have not proactively shared hazard and accident reporting information with Osprey units and unit safety personnel in the other services that operate the aircraft. Determining a process to proactively share relevant safety information with these personnel would provide greater assurance that Osprey units have the information needed to update their safety procedures. Aircraft knowledge and emergency procedures. Program stakeholders did not convene a multi-service conference or other forum to share Osprey aircraft knowledge and emergency procedures for 5 years (from 2020 to 2025). Service-specific changes to operational practices included safety related information, but these changes were not readily shared among the services, according to unit personnel with whom GAO spoke. The military services that operate the aircraft held a conference in May 2025, but they had not formalized plans to continue to do so. Without such a routine method, Osprey units have missed out on opportunities to share information that would enhance the safe operations of the aircraft. Maintenance data for common aircraft components and parts. Program stakeholders have taken steps to address maintenance data integrity issues for the hundreds of common Osprey aircraft components and parts that are shared across the services, but they have yet to confirm that all implementation steps have been completed. Without conducting a comprehensive review of Osprey maintenance guidance and inspection procedures, program stakeholders do not have assurance that efforts to improve maintenance information sharing have resolved data integrity issues, including for critical life-limited Osprey components, which has hindered their ability to ensure the safe operation of the aircraft. Why GAO Did This Study This testimony summarizes the information contained in GAO's December 2025 report, entitled Osprey Aircraft: Additional Oversight and Information Sharing Would Improve Safety Effort (GAO-26-107285). For more information, contact Diana Moldafsky at moldafskyd@gao.gov.

What GAO Found To understand, prevent, and treat infectious diseases, researchers study biological agents, such as bacteria and viruses. In the U.S., federal agencies have established guidelines to help ensure that U.S. biomedical research labs minimize biosafety and biosecurity risks. Certain principles or “key components” of biosafety and biosecurity may help reduce risks of all biological agents and research. GAO identified 10 key components that describe key steps a U.S. lab should take to mitigate the risks of biological agent research. The comparability of the selected Group of Twenty (G20) members’ relevant guidance documents to the 10 U.S. key components varied widely. Nine of the 10 selected G20 members in GAO’s review had documents that were comparable to one or more of the U.S key components for all biological agents and research. The U.S. key components include additional precautions for specified high-risk agents—such as Ebola virus—and research. Guidance documents from Australia, Canada, and China included comparable language to most of the additional precautions GAO identified for U.S. key components of biosafety and biosecurity. National guidance documents addressing biosafety and biosecurity are important, but other factors might also influence a G20 member’s biosafety and biosecurity. For example, Australian officials told GAO that state and territory governments play a role in managing biosecurity, such as responding to animal disease outbreaks. Why GAO Did This Study Governments use laws, regulations, policies, and guidelines to help achieve goals, such as protecting public health and safety. Biosafety helps protect lab workers, the community, and the environment from accidental exposure to or release of biological agents. Biosecurity helps protect against the loss, theft, deliberate release, or misuse of biological agents. The CARES Act includes a provision for GAO to monitor federal efforts in response to the COVID-19 pandemic. GAO was also asked to compare the biosafety and biosecurity standards of G20 members with U.S. standards. This report examines the extent to which selected G20 members’ publicly available guidance documents reflect (1) selected key components of U.S. biosafety and biosecurity for all biological agents and research and (2) additional precautions of the U.S. biosafety and biosecurity key components specific to high-risk biological agents and research. GAO analyzed core U.S. biosafety and biosecurity documents to identify 10 selected key components of U.S. biosafety and biosecurity for biomedical research labs. For each key component, GAO identified subcomponents or additional precautions for high-risk agents and research. GAO analyzed selected G20 members’ publicly available guidance documents, such as national laws, regulations, policies, and guidelines. GAO examined whether members’ documents had components that were comparable, somewhat comparable, or not comparable to the 10 U.S. key components. GAO did not evaluate the extent to which each member implements or enforces these documents. For more information, contact Karen L. Howard, PhD at HowardK@gao.gov.

What GAO Found Over the past decade, housing finance increasingly relied on nonbank mortgage companies (nondepository institutions specializing in mortgage lending). Nonbanks now originate and service most loans in the over $9 trillion in securities guaranteed by Ginnie Mae, a government-owned corporation, and by Fannie Mae and Freddie Mac, which are under Federal Housing Finance Agency (FHFA) conservatorship. From 2014 to 2024, the share of such loans serviced by nonbanks grew from 27 percent to 66 percent. But nonbanks have certain risks, such as reliance on short-term credit that may become unavailable during economic downturns. The failure of a large nonbank—or multiple smaller ones—could disrupt mortgage markets and increase federal fiscal exposure. Percentage of Mortgage Loans in Federally Backed Securities Serviced by Nonbanks, by Dollar Volume, 2014–2024 Ginnie Mae and FHFA have processes to assess the financial condition of nonbanks, but opportunities exist to enhance their processes. Financial data. Both agencies analyze nonbanks’ self-reported financial data to support their nonbank monitoring. However, GAO found that FHFA does not have written procedures to assess the reliability of these data, reducing assurance that its analytical results are dependable. Watch lists. Both agencies produce watch lists of nonbanks that pose relatively higher risks, based partly on financial data. But, based on GAO’s analysis, both agencies’ processes do not fully assess key risks of certain short-term credit lines. As a result, the agencies may not be fully considering information material to watch list determinations. Scenario analyses. Both agencies analyze the effect of changing economic conditions on nonbank financial health. To help manage its counterparty risk from nonbanks, Ginnie Mae performs a detailed analysis and uses a comprehensive economic stress scenario. But Ginnie Mae’s focus on a single adverse scenario does not reflect a fuller range of possible outcomes—potentially diminishing its ability to prepare for how different scenarios could affect its portfolio and financial exposure. Why GAO Did This Study Nonbank mortgage companies generally do not have a federal regulator overseeing their safety and soundness. But Ginnie Mae and FHFA, which support the stability of markets for mortgage-backed securities, play a role in monitoring these entities. Since 2013, GAO has designated the federal role in housing finance as a high-risk area. This report examines the (1) role of nonbanks in the mortgage market since 2014, including their benefits and risks; and (2) extent to which Ginnie Mae and FHFA designed processes to assess the financial condition of nonbank mortgage companies. GAO reviewed Ginnie Mae and FHFA policies, procedures, and analysis of nonbanks. GAO analyzed industry and government data on mortgage loans originated or serviced by nonbanks in 2014–2024. GAO also interviewed FHFA, Ginnie Mae, and Financial Stability Oversight Council officials, as well as researchers and subject matter experts from industry and consumer groups.

What GAO Found As of June 2025, the Internal Revenue Service (IRS) processed nearly 5 million Employee Retention Credit (ERC) claims. IRS moved quickly to administer ERC but was less prepared to assess improper payment risks and process a surge in claims. To address improper claims, IRS implemented a processing moratorium in September 2023. IRS closed most claims by December 31, 2025, according to IRS officials. GAO identified six lessons from ERC design and administration. Lessons Learned from the Design and Administration of Employee Retention Credit Offering Relief Through Employment Taxes Provides Benefits and Challenges Benefits include availability to employers without tax liability. Challenges include interactions with income tax. Some Design Decisions Increased Complexity and Improper Payment Risk Complex and retroactive eligibility criteria complicated eligibility determination. The Internal Revenue Service (IRS) Would Have Benefitted from a Comprehensive Plan for Managing Employee Retention Credit (ERC) Risks Timely implementing a 2022 GAO recommendation on project planning could have better prepared IRS for a later surge in claims. IRS Would Have Benefitted from Additional Eligibility Reporting Key eligibility information was not required on employment tax returns. Manual Processing for Amended Returns Complicated Compliance Efforts Paper-only amended returns limited IRS’s ability to capture key data. ERC Implementation Could Have Benefitted from More Timely and Consistent Communication with Stakeholders. IRS did not regularly communicate status of ERC processing. Source: GAO. | GAO-26-107456 These lessons could help policymakers consider future emergency employment tax relief, and help IRS better prepare for it. IRS did not complete an improper payment estimate for ERC, as required in law. The Department of the Treasury said it would not do so for pandemic programs as they are short term. However, a timely estimate could have helped identify root causes of improper payments earlier and developing one now could guide future decisions on employment tax relief. The statute of limitations for assessing tax on certain paid improper ERCs has expired. However, IRS can still pursue fraud cases indefinitely. Employers primarily claimed ERC on paper amended returns. IRS enabled electronic filing in mid-2024 but continued to process the returns manually. Automated processing would yield cost savings and expedite refunds. IRS’s last public update on ERC processing status was in October 2024, leaving uncertainties about cash flow among some employers. IRS also did not follow all risk management and internal control principles from GAO’s A Framework for Managing Improper Payments in Emergency Assistance Programs. IRS could reduce future improper payments by incorporating this framework into its policies. As a consequence of its design and administrative challenges, most ERC claims were not paid in 2020 or 2021, the eligibility period for the credit. About 83 percent of ERC refunds—about $235 billion—were issued in 2022 through June 2025, well after unemployment had returned to its pre-pandemic level. Why GAO Did This Study The ERC—which encouraged employers to keep paying employees during the COVID-19 pandemic—had provided about $283 billion to employers as of June 2025. GAO previously found that implementing new initiatives—such as the ERC—is a challenge for IRS. A law passed in July 2025 affected ERC by, in part, disallowing certain unpaid claims made after January 31, 2024. In response to a request, this report presents lessons learned on the ERC’s design and administration, examines actions IRS can take to be better prepared for emergency employment tax relief, and describes economic conditions—such as unemployment levels—surrounding ERC. To identify lessons learned, GAO reviewed literature and interviewed experts and agency officials about the ERC’s design and implementation. GAO observed ERC processing at an IRS campus. GAO compared documents with selected practices for managing payments in emergency programs (GAO-23-105876). GAO also analyzed IRS data on ERC processing and compared it with economic data.

What GAO Found Since February 2025, the Consumer Financial Protection Bureau (CFPB) has taken actions to reduce the size and scope of its activities and staffing. These actions included issuing stop-work orders; closing supervisory examinations; and terminating employees, contracts, and enforcement cases. Some of these actions, such as employee termination, are the subject of ongoing litigation and have not been finalized. On August 15, the United States Court of Appeals for the District of Columbia Circuit vacated the district court's injunction that prevented CFPB from taking various personnel actions, terminating certain contracts, and other actions related to downsizing the agency. The circuit court delayed the effective date of its order to allow the plaintiffs time to request a rehearing. Why GAO Did This Study CFPB's statutory duties include enforcing compliance with federal consumer financial laws, handling consumer complaints, promoting financial education, and monitoring financial products markets to identify risks to consumers. According to CFPB's acting leadership, in response to executive orders, the agency has been assessing ways to fulfill its statutory duties as a smaller, more efficient operation. GAO was asked to review the effect of recent stop-work orders, workforce reductions, contract terminations, and other related actions on CFPB's ability to fulfill its statutorily mandated functions. This report describes the status of CFPB's significant reorganization efforts as of August 2025. GAO will examine the effects of these actions as the subject of a future report. GAO reviewed publicly available information on CFPB's reorganization efforts, including court filings involving CFPB, Federal Register notices, press releases, executive orders, memorandums from the Office of Personnel Management and Office of Management and Budget, and nonpublic CFPB documents regarding detailees to the agency.  CFPB expressed concerns with the accuracy of this report. GAO stands by the accuracy of the facts presented, as discussed in the report. For more information, contact Alicia Puente Cackley at cackleya@gao.gov.