What GAO Found The National Science Foundation (NSF) has 21 research infrastructure projects funded through its Major Research Equipment and Facilities Construction (MREFC) and Research and Related Activities accounts, as of July 2025. This includes 13 major projects ($100 million or more) and eight midscale projects ($20 million to $100 million) at various stages of design, construction, and implementation. While all of these research infrastructure projects remained within their NSF-authorized total cost since GAO’s June 2024 report, several have experienced schedule delays or scope changes. Specifically, as of July 2025, four of the seven major projects in construction reported delays of 4 to 27 months relative to schedules GAO reported in June 2024 (see table below). NSF attributed delays to numerous factors, such as labor shortages, contractor underperformance, and budgetary uncertainty. Further, NSF reported reductions in scope for two of these projects, as well as three of eight midscale projects. Status of NSF’s Major Research Infrastructure Projects Under Construction, as of July 2025 Project Authorized cost Estimated completion Scope reduction Antarctic Infrastructure Recapitalization Program $155.4 million August 2029   -   Vera C. Rubin Observatory $571 million January 2026 ▲ 10 months - Antarctic Infrastructure Modernization for Science $410.4 million May 2027 ▲ 4 months ✓ Regional Class Research Vessels $400 million April 2029 ▲ 27 months ✓   Large Hadron Collider High Luminosity Upgrade Program       ATLAS Detector $82.8 million December 2028 ▲ 5 months - CMS Detector $88 million June 2028 ▼ 1 month - Leadership-Class Computing Facility $457.4 million March 2028 - Total $2,165 million     Legend: ATLAS = A Toroidal Large Hadron Collider Apparatus; CMS = Compact Muon Solenoid; ▲ = increase since June 2024; ▼ = decrease since June 2024; ✓ = scope reduced since June 2024. Source: GAO analysis of National Science Foundation (NSF) information. | GAO-26-107842 NSF considers several factors when selecting and awarding new research infrastructure projects. Specifically, NSF examines the scale and maturity of the proposed project, the availability and stability of annual appropriations, and external economic conditions. NSF has two separate processes for selecting major and midscale projects. Major facilities undergo an extensive, multiphase review and selection process, including consultation with the National Science Board. Midscale projects undergo a merit review and selection process led by NSF. Why GAO Did This Study Modern and effective research infrastructure, including facilities and equipment, is critical to maintaining U.S. global leadership in science and engineering. NSF provides funding for the design, construction, and operations of this infrastructure. This infrastructure spans a wide range of projects, from oceanographic research vessels to telescopes and supercomputers. The Consolidated Appropriations Act, 2024 includes a provision for GAO to review projects funded from NSF's MREFC account. This is the eighth report in this series and builds on GAO’s previous work. This report describes (1) the cost and schedule performance for NSF’s major and midscale projects funded through the MREFC account and (2) NSF’s process for selecting which projects receive MREFC funding and the key factors that contribute to NSF’s ability to select new projects. GAO examined NSF policies and documents for projects that were in design, construction, and implementation and interviewed agency officials. For more information, contact Hilary Benedict at benedicth@gao.gov.

What GAO Found Since 2016, the Department of Defense (DOD) has been developing the National Background Investigation Services (NBIS)—an IT system for conducting background investigations for most federal agencies and over 13,000 industry organizations that work with the government. DOD originally expected NBIS to be complete in 2019, but repeated delays have hindered deployment. GAO has also found that the previous NBIS cost estimate and schedules were unreliable. After missing multiple targets, DOD’s Defense Counterintelligence and Security Agency (DCSA) paused NBIS development in 2024 to revise its approach. In 2025, it developed a new cost estimate and changed its approach to scheduling. GAO reviewed DCSA’s 2025 NBIS cost estimate and found it to be reliable because it at least substantially met four characteristics of a reliable cost estimate. DOD now projects spending an additional $2.2 billion on NBIS development through fiscal year 2031, which is in addition to $2.4 billion previously spent on NBIS and legacy systems through fiscal year 2024. A reliable cost estimate should help prevent unexpected cost overruns and provide needed visibility. GAO also found that DCSA continues to lack a reliable schedule for NBIS. The program’s schedule showed improvements, substantially meeting two characteristics of a reliable schedule. However, it only partially met the credible and well-constructed characteristics. For example, a risk analysis can help the program prioritize those risks that may lead to delays. Not having a reliable schedule is likely to continue to affect the NBIS program’s ability to meet milestones, including the goal to complete development in fiscal year 2027. Assessment of Department of Defense’s 2025 NBIS Cost Estimate and Schedule Against Best Practices Continued prioritization of NBIS development will be critical to successfully implement government-wide personnel vetting reforms and address persistent challenges. For example, GAO has found federal agencies have not met timeliness goals for nearly all phases of the security clearance process. In particular, average times for initial Top Secret clearances have consistently trended longer from fiscal year 2022 to 2025. Sustained leadership attention is key to fully deploying the NBIS system and achieving personnel vetting reform. Why GAO Did This Study U.S. government personnel vetting processes, such as background investigations, rely on IT systems to process data on millions of federal employees and contractor personnel. DOD has been developing NBIS as the new IT system for personnel vetting since 2016. In 2018, the government initiated a major reform of personnel vetting to better protect national security called Trusted Workforce 2.0. GAO placed the personnel vetting process on its High-Risk List in the same year. This statement discusses (1) the reliability of DCSA’s 2025 cost estimate for the NBIS program, (2) the extent to which the NBIS program has met scheduling best practices, and (3) the importance of NBIS to achieve personnel vetting reforms under Trusted Workforce 2.0. This statement is based on GAO’s analysis of DCSA’s 2025 cost estimate and schedule, prior GAO reports on NBIS from December 2021 through September 2025, and ongoing work assessing NBIS development. To perform prior and ongoing work, GAO analyzed information on NBIS from DCSA and interviewed knowledgeable officials.

What GAO Found Paid preparers have differing levels of skills, education, and expertise. Tax practitioners include enrolled agents, certified public accountants, attorneys, and other individuals who possess some level of qualification or credentials issued by either Internal Revenue Service (IRS) or states. All other paid preparers without professional credentials are considered unenrolled preparers. While all paid preparers are required to have and use a Preparer Tax Identification Number (PTIN) and are subject to various requirements in the Internal Revenue Code, IRS only has the authority to regulate the practice of tax practitioners. Generally, unenrolled preparers are not subject to IRS regulation, including testing and education requirements. Number and Type of Credentials Held by Paid Preparers for 2025 In prior work, GAO found that paid preparers can make serious errors on the tax returns they prepare. In addition, GAO’s prior work indicates that unenrolled preparers can make errors at a higher rate than taxpayers who prepare their own returns or who use other categories of paid preparers. When paid preparers make errors, taxpayers may overpay and lose out on tax benefits. Alternatively, when preparers understate tax liabilities, taxpayers may be subject to penalties and the government may collect less revenue. GAO has previously reported that IRS uses various tools, including outreach and education, civil and criminal investigations, and penalties to oversee all paid preparers and bring them into compliance. For example, IRS’s Refundable Credits Return Preparer Strategy program identifies preparers who were potentially noncompliant with due diligence requirements and encourages them to comply. Actions IRS may take range from issuing warning letters and phone calls to preparers to more serious actions such as audits of preparers’ clients and IRS staff visits to preparers. IRS conducts civil and criminal investigations of abusive tax schemes, including those involving paid preparers. Paid preparers may be subject to penalties for noncompliance with certain requirements in the Internal Revenue Code. Why GAO Did This Study During fiscal year 2024, more than half of all individual taxpayers used a paid preparer, according to IRS data. Paid preparer errors can lead to billions of dollars in improper claims of refundable tax credits and can have negative consequences for individual taxpayers. GAO was asked to examine IRS’s ability to oversee preparers. This report describes what GAO and other IRS oversight groups have previously reported on IRS’s oversight of paid preparers and recommended to improve its efforts. GAO reviewed the Internal Revenue Code, relevant regulations and case law, information on paid preparer credentials, GAO’s previous issued work on paid preparer topics, and reports by other IRS oversight groups. Additionally, GAO analyzed data on numbers of preparers with a current PTIN and interviewed IRS officials.

What GAO Found Since January 2025, the President has issued several directives for federal agencies to restrict hiring and reduce the size of the federal government’s workforce (see figure below for examples). However, agencies were still allowed to hire new staff for certain positions, including those related to national security and public safety. Timeline of Selected Directives and Government-wide Guidance on Federal Workforce Changes, January to June 2025 Data reported by major federal agencies showed that from January to June 2025, staffing declined at nearly all of them. Across these agencies, about 134,000 employees (or 6 percent of the workforce) separated during this period, while about 66,000 employees (including temporary employees) were hired. Agencies reported that another roughly 144,000 employees were approved for a deferred resignation program and would end their federal employment by the end of 2025. The Office of Personnel Management (OPM) has updated how it reports federal workforce data. OPM has broad responsibilities for collecting and sharing workforce data. In January 2026, the agency launched a website with new features and more frequent data releases to make workforce data more accessible and useful for Congress and others. Why GAO Did This Study GAO was asked to provide quarterly updates on workforce changes at the 24 federal Chief Financial Officers (CFO) Act agencies. This report (1) provides information on workforce changes at CFO Act agencies from January to June 2025, and (2) describes actions OPM has taken to modernize how it reports workforce data. To address these objectives, GAO collected data from CFO Act agencies on workforce changes during this period, and information from OPM on updates to how it collects and presents agency data. GAO will provide updates to the information in this product in future reports on these issues. For more information, contact Dawn G. Locke at LockeD@gao.gov.

What GAO Found Anyone conducting weather modification operations in the U.S. is required to notify and update the National Oceanic and Atmospheric Administration (NOAA), which is in turn required to maintain a record of the activities and share it with the public. Weather modification includes cloud seeding (which seeks to alter local precipitation) and solar geoengineering (which seeks larger scale changes in climate). Several states use cloud seeding to address an ongoing lack of precipitation in the western U.S. Cloud seeding is the most common weather modification technique, but recently, more organizations have begun to conduct solar geoengineering operations or research. NOAA has a statutory requirement to provide oversight of weather modification reporting, but the agency is not fully meeting its responsibilities to maintain and share weather modification reports. Weather modification operators may report inconsistent information or fail to report, and we estimate that over half of all the reports filed with NOAA likely have errors, including missing required information (e.g., maps). NOAA does not have written agency guidance for reviewing reports or maintaining its database. Such guidance could help NOAA meet its responsibilities and provide useful, complete, and transparent information to the public. Estimated Errors in Initial, Interim, and Final Reports from the National Oceanic and Atmospheric Administration (NOAA) Weather Modification Reports Database NOAA is also required to maintain a record of emerging solar geoengineering activities, but its forms and processes are not well-suited to those activities. Operators had challenges with the forms and reporting process and may also be unaware of the reporting requirement. State and local officials engaging with weather modification activities may likewise be unaware of NOAA’s reporting process, even though the reports could contain information they need to make decisions about weather modification. Improved instructions and outreach from NOAA could help ensure operators are reporting their activities and providing better-quality information to improve understanding of this emerging technology. Implementing measures to increase the quality of the information NOAA collects from operators could help provide transparent information to the public about weather modification activities, including solar geoengineering. This could also provide more assurance that these activities are being properly overseen, and may help address increased public concern, canceled projects, or bans. Why GAO Did This Study NOAA, within the Department of Commerce, is the only federal agency with responsibilities under the Weather Modification Reporting Act of 1972 (the Act). This report examines the extent to which NOAA is meeting its statutory responsibilities for overseeing current U.S. weather modification reporting and is prepared to oversee an emerging approach like solar geoengineering. To conduct this review, GAO analyzed a representative sample of weather modification reports filed with NOAA and available on its website and compared reports to information from other sources. GAO reviewed 54 articles from a structured literature search, visited nine locations across two states, and held 24 interviews with agency officials, operators, and other stakeholders.

What GAO Found The U.S. Department of Health and Human Services (HHS) leads the federal public health and medical preparedness for, response to, and recovery from public health emergencies. HHS awarded almost $900 million in fiscal year 2024 to states and other jurisdictions to help them prepare for a range of public health threats and emergencies through two programs. First, the Public Health Emergency Preparedness (PHEP) program supports jurisdictions’ public health systems, which includes public health agencies at all government levels. Second, the Hospital Preparedness Program (HPP) supports health care systems, which include a community’s health care organizations such as hospitals. Jurisdictions distribute HPP funds to health care coalitions—made up of health care and other entities to support preparedness. HHS requires jurisdictions and health care coalitions to complete activities. These include exercises to help ensure that jurisdictions are prepared to respond to public health threats, such as infectious diseases and extreme weather events (e.g., hurricanes). Debris from Damaged Homes Following Hurricanes Helene and Milton, 2024, Florida The two HHS agencies that administer these programs lack a formal mechanism, such as joint exercises, written agreements, or working groups, to coordinate them. Coordinating these preparedness programs could allow HHS to better manage them and support jurisdictions as they prepare both their public health and health care systems to respond to public health threats and emergencies. Lessons learned from the COVID-19 pandemic have demonstrated the importance of coordination between these systems, such as between state health departments and hospitals, if jurisdictions are to be effectively prepared. Additionally, officials from selected jurisdictions said that greater interagency coordination could help reduce resource inefficiencies associated with implementation of PHEP and HPP. Further, HHS does not collect or analyze information on jurisdictions’ ability to meet the 15 public health and four health care preparedness capabilities and any related gaps. According to HHS documentation, it identified these capabilities to serve as national guidance. The capabilities describe skills and abilities jurisdictions need to effectively respond to, and recover from, public health threats and emergencies. For example, they include providing mass medical care and laboratory testing at scale for emerging pathogens. Collecting and analyzing such information would help HHS understand the extent to which jurisdictions are prepared to respond to public health threats and emergencies and whether any changes are needed. Why GAO Did This Study The United States faces ongoing risks from future public health threats, such as infectious diseases; extreme weather events; and biological, chemical, nuclear, and radiological events. In fiscal year 2024, HHS, through PHEP, provided $654 million, and through HPP, provided $240 million, to jurisdictions to aid preparedness for such threats. The CARES Act includes a provision for GAO to examine public health preparedness and response efforts related to the COVID-19 pandemic. This report examines, among other objectives, the extent to which HHS agencies (1) coordinate PHEP and HPP and (2) collect and analyze information on jurisdictions’ capabilities. GAO reviewed HHS documentation, including notices of funding opportunity, templates, and examples of documents submitted by eight selected jurisdictions. GAO also interviewed officials from HHS and the eight selected jurisdictions. GAO selected these jurisdictions to include variation among geographic location, percent of the population living in a rural setting, and public health governance structure.

What GAO Found Natural disasters at military installations have resulted in significant costs and damages since 2015, but the Department of Defense (DOD) has not comprehensively tracked data related to those costs. In 2024, DOD began an effort to track the effects of extreme weather at military installations. However, GAO identified gaps in the scope of DOD’s data collection and its ability to collect complete and accurate data on disaster recovery costs. Specifically, DOD’s data collection is limited to the effects of extreme weather, which does not include the full scope of natural disasters, including seismic events such as earthquakes. Also, data DOD collects on the cost of extreme weather at installations may be inaccurate or incomplete in some cases, in part due to the timing of when installations are expected to report the information. Expanding the scope of its data collection to include all types of natural disasters and establishing a process to ensure cost data are complete and accurate can improve DOD’s ability to anticipate future disaster recovery needs. Figure: Examples of Natural Disasters at Military Installations DOD has taken steps to increase installations’ disaster resilience—including implementing resilience improvements at the 12 installations GAO examined—but gaps in planning may limit these efforts. GAO identified some instances when installations were unable to pursue resilience improvements due to not having necessary data or sufficient funding. Installations are working to implement a statutory requirement to include resilience information as part of their master plans. This includes the identification of current and future risks and ongoing or planned projects to mitigate those risks. DOD policy reflects this requirement and addresses how resilience should be incorporated into construction projects, such as by adhering to Unified Facilities Criteria (DOD’s standards for design and construction of facilities). However, existing DOD and military department guidance does not address how installations should use the master plan resilience information when recovering from a disaster. By including such information in guidance, the military departments can help ensure that installations affected by disasters are better able to incorporate resilience improvements while quickly restoring essential capabilities. Why GAO Did This Study Extreme weather and natural disasters have resulted in billions of dollars of damages to military installations over the past decade. Such damages can affect DOD’s ability to execute its mission and disrupt installation resources that support service members and their families. Senate Report 118-58, accompanying a bill for the National Defense Authorization Act for Fiscal Year 2024, includes a provision for GAO to assess DOD’s recovery from natural disasters at military installations. GAO examined the extent to which DOD has (1) tracked costs and damages associated with natural disasters since 2015 and (2) increased resilience of selected installations affected by natural disasters. GAO reviewed documentation and interviewed officials from a non-generalizable sample of 12 installations on disaster damages and resilience improvement efforts and conducted in-person site visits at three of these installations. GAO analyzed annual and supplemental appropriations to identify funds designated for disaster recovery. GAO also reviewed documentation and interviewed DOD and military department officials related to installations’ recovery efforts and collection of related data.

What GAO Found The Consumer Product Safety Commission (CPSC) uses risk-based approaches to target children’s products that may contain toxic substances for examination at U.S. ports. For example, it uses U.S. Customs and Border Protection systems that analyze national shipment and law enforcement data to target high-risk products. Examinations include screening products with handheld devices for lead and other toxic substances and reviewing importer documentation to verify that products were tested by third-party labs and meet CPSC safety standards. CPSC’s Process for Examining Children’s Products at U.S. Ports Starting July 2026, CPSC will require importers to electronically submit (“e-file”) key data (including product identification and place of testing) when products enter U.S. ports. According to CPSC officials, e-filing may help address challenges, such as delays in completing examinations when importers lack lab testing documentation. However, CPSC has not developed an oversight plan to ensure that importers file timely, accurate data. Establishing such a plan would help CPSC ensure that e-filing achieves its intended objectives, such as making targeting more effective and examinations more efficient. CPSC has review processes to verify that third-party labs that test children’s products meet its accreditation and other requirements. For example, labs owned or controlled by a manufacturer or government entity must provide information about their safeguards from undue influence. However, CPSC has not proactively analyzed data across all types of labs to assess potential risks, such as inaccurate testing or misreported results. CPSC recently began analyzing violations data for labs owned by manufacturers to better evaluate their safeguards from undue influence, but it does not do so for independent or government labs. By better leveraging its violations data for these labs, CPSC would be better positioned to identify and address potential problems associated with all types of labs, which could help prevent violative products from entering the market. CPSC has reviewed and updated some testing requirements for children’s products, but not its requirements for lead or phthalates (toxic substances used to make plastics more pliable). The Consumer Product Safety Improvement Act of 2008 requires CPSC to review its lead requirements at least every 5 years; however, it has not done so. In addition, CPSC does not have written procedures for how staff should monitor changes related to phthalates and other toxic substances in children’s products. By reviewing its lead requirements and documenting a process for staying up-to-date on toxic substances, CPSC could help ensure that it does not miss opportunities to strengthen its standards and protect children from harm. Why GAO Did This Study CPSC is responsible for regulating the safety of thousands of consumer products, including children’s products. It requires manufacturers and importers to have certain toys and other children’s products tested by labs for lead and other toxic substances before they can enter the U.S. market. However, the large volume of products entering U.S. ports makes it challenging to ensure compliance. GAO was asked to review how CPSC addresses children’s products that contain toxic substances. This report examines CPSC’s (1) processes for examining children’s products and planning for an electronic data filing system; (2) processes for approving and assessing risks of third-party labs that test children’s products; and (3) efforts to keep safety standards for lead and other toxic substances up-to-date. GAO reviewed CPSC’s third-party testing and lab accreditation requirements; analyzed CPSC policies and procedures; and visited a nongeneralizable sample of four ports to observe product examinations. Ports were selected to prioritize those with higher volumes of children’s products entering the U.S. and to provide a mix of geographic locations.

What GAO Found Seven percent of participants in the Department of Labor (DOL) workforce programs GAO reviewed reported having disabilities, according to DOL data from July 1, 2023 through June 30, 2024. Participants with disabilities had worse employment outcomes. Of those reporting a disability, 43 percent did not report their disability type (e.g., a physical or cognitive disability). GAO found that the completeness of these data varied widely by state. DOL has taken steps to improve the completeness of these state-collected data, which could help it improve outcomes for participants with different types of disabilities. However, officials said they do not plan to target assistance to states with less complete data. State and local workforce agencies GAO visited have made various efforts to ensure their programs are accessible to participants with disabilities. For example, officials GAO interviewed said they have provided assistive technology such as computers that are accessible to people with visual impairments (see figure). However, officials also reported challenges ensuring that training provided outside of job centers is accessible and providing certain accommodations such as sign language interpreters. Computer Workstation with Accessibility Features at a Job Center DOL monitors state and local workforce agencies to help ensure its programs are accessible but does not routinely analyze monitoring results. DOL’s monitoring reports include an examination of state and local accessibility efforts, but the agency has not developed a procedure to analyze them at an aggregate level, which increases the risk the agency will fail to identify and address widespread issues. Also, DOL has provided guidance and technical assistance on a range of accessibility topics. However, officials GAO interviewed in several local areas were not aware of or do not use these resources. Without evaluating awareness or use of its guidance and assistance, DOL may miss opportunities to improve dissemination of materials or take other actions to promote them. Why GAO Did This Study The Workforce Innovation and Opportunity Act authorizes programs that help job seekers access services at locations called job centers. It prohibits discrimination, requiring that services be accessible to people with disabilities. GAO was asked to examine efforts to ensure programs are accessible. This report addresses (1) the share of participants in selected DOL workforce programs that report disabilities and use services, and their employment outcomes, (2) the efforts made and challenges faced by state and local workforce agencies to ensure the programs’ services are accessible to people with disabilities, and (3) DOL’s efforts to ensure the programs are accessible. GAO analyzed participant-level DOL data from July 1, 2023 through June 30, 2024, the most recent year available; reviewed relevant federal laws, policies, and agency documents; interviewed DOL and state officials, disability organizations, and participants with disabilities; and visited a non-generalizable sample of 12 job centers in Arkansas, Oregon, Virginia, and the District of Columbia, selected for variation in geographic location, urban or rural population, and percentage of participants with disabilities.

What GAO Found The Department of the Interior has a goal to improve data sharing, which is vital to its oversight of oil and gas development on federal and Tribal lands. Three bureaus within Interior share oversight responsibilities—the Bureau of Indian Affairs (BIA), Bureau of Land Management (BLM), and Office of Natural Resources Revenue. Interior and the bureaus have reported spending about $206 million since 2021 to modernize data systems. However, GAO identified challenges in three selected business processes—leasing land for oil and gas exploration, verifying that production is reported accurately, and reviewing bonds to ensure they are sufficient to cover cleanup costs. For example: Manual data sharing. To update data on oil and gas leases, BIA receives lease data in paper documents from companies that it then manually enters into its data system. BIA then emails or mails these data to other Interior bureaus, whose staff must manually enter the data into their systems. Inability to accept electronic data. To help ensure accurate royalty payments for extraction of natural gas, BLM verifies selected gas analysis reports submitted by operators. Since 2019, federal regulations have required operators to submit these reports electronically to BLM’s Gas Analysis Reporting and Verification System. However, this system has never been operational, and BLM has discontinued estimating when it will be. As a result, BLM manually collects data from operators as needed and still does not have a system to receive and store reports and data automatically. Oil and Gas Data-Sharing Challenges Identified at the Department of the Interior Business process Bureau of Indian Affairs Bureau of Land Management Office of Natural Resources Revenue Leasing • Manual data sharing • Manual data sharing • Manual data sharing Production verification N/A • Manual data sharing • Data not accepted by systems No challenges identified Reviewing bonds • Data not accepted by systems • Manual data sharing N/A Legend: N/A = not applicable. The bureau does not perform this business process. Source: GAO analysis of Interior data systems. | GAO-26-106475 Note: Some automated data sharing may also be used for other aspects of these business processes. Interior issued a 2023 data-sharing policy that targeted further enhancements across the department. However, this strategy does not have time frames and measurable performance goals for improving data sharing. Accordingly, Interior does not know when needed improvements will materialize. Without enhanced automated data sharing, highly inefficient and error-prone manual processes will continue unabated. Interior and its oil and gas bureaus have generally followed internal data governance policies but have not completed data-sharing agreements as required by Interior policy. These agreements are supposed to guide how officials use and manage shared data. They also could help officials address the inherent challenges of manual processing. Without such agreements, Interior is likely missing opportunities to automate and electronically share information. Why GAO Did This Study Development of oil and gas resources on federal lands helps supply the U.S. with energy and generates billions of dollars annually in revenues. Within Interior, selected bureaus use data systems to oversee oil and gas development, such as by issuing drilling permits and inspecting wells. GAO was asked to review Interior’s oil and gas data systems. This report (1) describes data sharing challenges, (2) examines Interior’s plans for improving data sharing, and (3) examines the extent to which Interior followed policies for data sharing. GAO reviewed three selected oil and gas business processes that Interior uses. GAO examined data system documentation and interviewed officials who carry out these three business processes to identify challenges. GAO reviewed agency data-sharing policies, documentation, and strategy documents and interviewed system administrators and agency leadership.

What GAO Found In February 2026, GAO identified five open recommendations under the purview of the National Science Foundation (NSF) Chief Information Officer (CIO), from previously issued work. Each of these recommendations relate to the Improving IT Acquisitions and Management GAO High-Risk area. For example, GAO previously recommended that NSF develop guidance regarding standardizing cloud service-level agreements. Further, GAO recommended that NSF complete annual reviews of its IT portfolio consistent with federal requirements. The CIO's continued attention to these recommendations will help ensure the effective use of IT at the agency. Why GAO Did This Study CIO open recommendations are outstanding GAO recommendations that warrant the attention of agency CIOs because their implementation could significantly improve government IT operations by securing IT systems, identifying cost savings, improving major government programs, eliminating mismanagement of IT programs and processes, or ensuring that IT programs comply with laws, among others. For more information, contact Nick Marinos at marinosn@gao.gov.

What GAO Found Research that involves modifying pathogens that have the potential to cause a pandemic—sometimes referred to as “gain-of-function research of concern”—has been a topic of debate. Based on GAO’s review of literature and other sources, this research has advanced scientific knowledge of how pathogens infect humans and transmit and cause disease. However, there is no broad agreement on the extent to which this research has directly led to the development of vaccines and therapeutics, such as for COVID-19. There was broad consensus that gain-of-function research of concern can pose biosafety and biosecurity risks. This is because this research can involve enhancing the transmissibility or virulence of pathogens that have the potential to cause widespread and uncontrollable disease, resulting in significant morbidity and mortality if they were to be accidentally or deliberately released from a lab. Scientist Conducting Pathogen Research As part of its effort to lead the federal public health and medical response to potential biological threats and emerging infectious diseases, the Department of Health and Human Services (HHS) provides funding for, and conducts research on, pathogens of varying risk level. GAO found that HHS procedures for reviewing research—including research that can be considered gain-of-function research of concern—generally include identifying and assessing the risks of the pathogen and the proposed experiment and assessing the adequacy and appropriateness of proposed risk mitigation strategies. If risks cannot be mitigated, HHS agencies can decide not to fund or conduct the research. However, GAO also found that HHS does not always share key information on these risk reviews with the public. For example, HHS reports to federal stakeholders about the number of research projects involving certain higher risk pathogens and the related risks and associated mitigation measures but does not report more widely. Some HHS officials told GAO they supported sharing general information about their risk reviews with the public. HHS has also reported that transparency helps to ensure public trust in federally funded scientific research. Sharing such information would help provide greater assurance to the public, science community, and Congress that HHS has procedures to manage risks. Why GAO Did This Study Recently introduced legislation and executive actions have aimed to restrict or ban federal departments and agencies, like HHS, from conducting or funding gain-of-function research of concern. GAO was asked to review the outcomes of gain-of-function research of concern and related risk mitigation strategies. This report (1) describes findings from literature and reports that discuss outcomes of gain-of-function research of concern and (2) examines HHS’s procedures for reviewing risk and risk mitigation strategies for research involving pathogens. GAO identified outcomes of gain-of-function research of concern by reviewing literature and other sources published from 2019 to 2024. GAO reviewed HHS procedures for reviewing risks and risk mitigation strategies and federal policies and guidance for the oversight of higher-risk pathogen research and interviewed HHS officials. GAO also interviewed eight biosafety and biosecurity experts selected because they authored relevant articles and had experience with gain-of-function research of concern.

What GAO Found The No Surprises Act, among its provisions, generally prohibits providers from balance billing in certain circumstances—such as emergency services—for individuals with private health insurance. Balance billing is when insured patients receive a bill from an out-of-network health care provider for the amount above any applicable cost-sharing that exceeds the health plan or issuer’s payment. An unexpected balance bill is referred to as a surprise bill. GAO analyzed the percentage of claims that were in-network for selected specialties to examine potential changes in network participation after the act’s implementation. Increases in the percentage of in-network claims may indicate increases in provider participation, while decreases may indicate reduced participation. Among specialties likely to be affected by the No Suprises Act protections—emergency medicine, radiology, anesthesiology, and air ambulance—the percentage of in-network claims increased for three of the four specialties after the act took effect. For example, GAO found the percentage of in-network facility claims (typically submitted by hospitals) and professional claims (typically submitted by physicians) for emergency medicine declined before the No Surprises Act took effect, then increased afterward. Percentage of In-Network Facility (Hospital) and Professional (Physician) Claims for Emergency Medicine, 2019–2023 Payment changes for the selected services largely reflected continuations of trends prior to the No Surprises Act taking effect. For example, the inflation-adjusted payment for in-network emergency medicine services billed by facilities increased in 2022 and 2023, continuing the trend since 2019. Meanwhile, the inflation-adjusted payment for in-network emergency services billed by physicians or their practices decreased in 2022 and 2023, continuing previous trends. Why GAO Did This Study Health plans or issuers contract with certain providers by negotiating payment rates to create provider networks that serve their patients. In-network providers accept negotiated payment rates (including any applicable patient financial responsibility) as full payment. Providers outside of that network do not have such contracts and have not agreed to a payment rate with the issuer. This can cause financial strain for patients if the providers bill them for the charges exceeding the issuer’s payment. The No Surprises Act was enacted on December 27, 2020. The act’s protections against surprise bills took effect beginning January 1, 2022. The Consolidated Appropriations Act, 2021, includes a provision for GAO to review the effects of the No Surprises Act. This report describes (1) changes in network participation, as measured by the percentage of in-network claims billed, for selected provider specialties before and after the No Surprises Act went into effect and (2) changes in payments for selected services before and after the No Surprises Act went into effect. GAO analyzed a large dataset that included health insurance claims, specifically looking at four specialties most likely to be affected by the No Surprises Act’s prohibition on balance billing. GAO examined trends in in-network claims and payments from 2019 through 2023. GAO also interviewed representatives from 20 stakeholder groups, including specialty provider associations, issuers, and state insurance departments, to discuss changes in network participation and payment rates after the No Surprises Act took effect. The Department of Health and Human Services and Department of Labor provided technical comments, which we incorporated as appropriate. For more information, contact John E. Dicken at dickenj@gao.gov.

What GAO Found Under the Bayh-Dole Act of 1980, federal agencies can, in certain circumstances, exercise the authority known as march-in rights when an invention that arose from federally funded research is involved. March-in-rights entail an agency requiring a recipient of its funding to issue a license to a third party to develop the invention. Agencies have never exercised march-in rights. In December 2023, the National Institute of Standards and Technology (NIST) published draft guidance that sought to clarify when agencies could exercise this authority. It proposed using the price of a product resulting from a federally funded invention as a factor for exercising march-in rights. According to the guidance, price could be used under two of the four statutory criteria: practical application and health or safety need (see figure). The draft guidance was developed through a NIST-led interagency process. As of December 2025, NIST did not have a timeline for finalizing the guidance, citing a lack of interagency consensus. NIST Draft Guidance Proposed Using Price as a Factor Under Two Statutory Criteria for Exercising March-In Rights Among the 51,762 public comments on the draft guidance, more than 47,000 comments (about 91 percent) expressed support for the draft guidance, with the remainder expressing opposition. Most comments in favor of the guidance expressed concern about high prescription drug prices and support for using march-in rights to lower them. Comments opposing the guidance—including all comments submitted by universities—raised concerns about potential adverse effects, such as reducing universities’ ability to license inventions and businesses’ ability to attract investment to develop the inventions into products. Because march-in rights have never been exercised, it is only possible to discuss hypothetical impacts of implementing the draft guidance. A federal agency could exercise march-in rights based on product price only if a product resulting from a federally funded invention has an unexpired patent subject to Bayh-Dole. Therefore, the potential for march-in is higher for technologies with a high volume of patenting activity arising from federally funded research, such as pharmaceuticals, computer technology, and electrical machinery. Although most public comments on the draft guidance expressed support for using march-in rights to lower drug prices, studies estimate that march-in based on price would likely affect a small number of drugs. This is because most drugs have patents that are not subject to Bayh-Dole. Why GAO Did This Study Federal agencies fund universities and other organizations to conduct research, which can lead to new inventions. Under the Bayh-Dole Act, recipients of federal funding can retain patent rights to the inventions and license them to other parties. To protect public interest in these inventions, the act allows federal agencies to retain certain rights, including march-in rights. These permit an agency to require a recipient to issue a license to a third party, when the circumstances meet at least one of four criteria specified in the act. If the recipient refuses, the agency itself can grant a license. Agencies can initiate march-in proceedings on their own or in response to requests from external parties. Since the passage of the Bayh-Dole Act, agencies have received about a dozen march-in requests; most of these addressed lowering the price of drugs or other medical technologies. For all the requests, agencies declined to exercise march-in rights. GAO was asked to review development of NIST’s draft guidance and its potential impacts. This report examines: (1) key elements of the draft guidance and the NIST-led interagency process for developing it; (2) stakeholder views on the draft guidance, as reflected in public comments; and (3) available information about the potential impacts of exercising march-in rights based on price. GAO reviewed applicable laws and regulations, analyzed public comments and patent data, reviewed studies estimating how many drugs could be affected by exercising march-in rights based on price, and interviewed agency officials. For more information, contact Candice N. Wright at wrightc@gao.gov.

What GAO Found With few exceptions, in states subject to Public Law 280 (P.L. 280), Tribes share criminal jurisdiction with the state, rather than the federal government, when a tribal citizen commits a crime or, in certain circumstances, is the victim of a crime in Indian country. However, P.L. 280 currently has limited effect on criminal jurisdiction in Alaska—a mandatory P.L. 280 state—because the Alaska Native Claims Settlement Act of 1971 eliminated most of the state’s Indian country. Nevertheless, Tribes in Alaska have inherent authority to exercise certain criminal jurisdiction which, due to other laws, is concurrent with the state. While Alaska and Department of Justice (DOJ) officials had consistent views, officials from the Department of the Interior’s Bureau of Indian Affairs (BIA) shared varying views on this concurrent jurisdiction during GAO’s review. By documenting BIA’s position on concurrent criminal jurisdiction in Alaska and making it readily available, BIA may help ensure consistent understanding of the authority of Tribes in Alaska. Alaska’s status as a mandatory P.L. 280 state may affect Tribes’ access to certain federal public safety and justice funding. Specifically, BIA data for fiscal years 2017 through 2021 (the most recent data available) showed that Tribes in Alaska and other mandatory P.L. 280 states received less funding than Tribes in all other states in the lower 48. Officials stated that BIA has limited discretionary funding and generally must continue to fund Tribes with existing funding agreements for public safety and justice programs at the same level as the prior year. When new discretionary funding is available, BIA officials stated they generally prioritize this funding for Tribes that do not receive public safety and justice services from their respective states. However, BIA has not documented criteria for entering into new funding agreements with Tribes for the first time. By doing so, BIA could provide greater transparency on funding decisions, particularly if Congress appropriates additional funding. BIA Public Safety and Justice Estimated Needs and Expenditures to Federally Recognized Tribes and Tribal Organizations, Fiscal Year 2017 Through 2021 Tribes by P.L. 280 status Tribal needs estimate BIA expenditures Alaska (Mandatory P.L. 280) $3,046,624,000 $59,519,000 Mandatory P.L. 280 states in the lower 48 $2,827,096,000 $151,738,000 All other states in the lower 48 $8,695,641,000 $1,864,373,000 Source: GAO analysis of Bureau of Indian Affairs (BIA) Tribal Law and Order Act reports for fiscal years 2017 through 2021. | GAO-26-107533 Tribal leaders, Alaska Native community members, as well as federal and state officials identified numerous challenges related to public safety and justice in Alaska Native communities. In general, these challenges included high crime rates and limited public safety infrastructure, funding resources, and training. Federal, state, and tribal officials identified various efforts to improve public safety and justice in Alaska, including state-tribal wellness courts. Why GAO Did This Study In 1953, P.L. 280 rescinded most federal criminal jurisdiction in Indian country in certain states and provided it to these select states. As of January 2026, this affects approximately 65 percent of the 575 federally recognized Tribes, including Alaska—a state with one of the highest crime rates in the U.S.. Following the passage of subsequent federal legislation, some uncertainty has been raised about the extent to which P.L. 280 is applicable in the state of Alaska. GAO was asked to review P.L. 280’s effect on public safety and justice on Alaska Native communities.This report addresses (1) the extent to which P.L. 280 affects criminal jurisdiction in Alaska; (2) the effects of P.L.280 on Tribes in Alaska; and (3) views on public safety and criminal justice challenges in Alaska. Among other methods, GAO analyzed relevant laws and federal memoranda on tribal jurisdiction in Alaska, as well as federal crime and funding data for fiscal years 2017 through 2024. GAO interviewed BIA and DOJ officials, as well as non-governmental stakeholders. GAO conducted listening sessions with a non-generalizable sample of over 50 tribal leaders and Alaska Native community members.

Our prior work on the Sentinel program and the Air Force’s planning for the transition from Minuteman III to Sentinel revealed multiple challenges and opportunities. Air Force’s Current and Future Intercontinental Ballistic Missile Systems Our key findings on the Sentinel program include the following: Opportunities during restructuring. As a result of the Nunn-McCurdy breach, the Under Secretary of Defense for Acquisition and Sustainment rescinded Sentinel’s Milestone B approval and related program baselines. Milestone B approval is a decision allowing major defense acquisition programs to enter into system development and demonstration prior to production and deployment. The Air Force is restructuring the program as it proceeds toward obtaining new Milestone B approval. The program’s first flight of the Sentinel missile has slipped about 4 years from original estimates and is now planned to occur in March 2028. However, the breach has provided the Air Force the opportunity to address fundamental issues with the program, including resolving issues with key design tools and performance requirements and finalizing the launch facility design. Future program outcomes will depend on the extent to which the Air Force takes advantage of this opportunity to correct earlier missteps. Major Events in the Sentinel System Development Options to lower costs and arrest delays. Sentinel program officials continue to evaluate options to potentially redesign portions of the weapon system for cost reductions and are looking at avenues to minimize further schedule delays. For example, the Air Force is reevaluating system requirements and evaluating changes to the acquisition strategy—both of which could limit further cost and schedule growth. Software risks. Sentinel is software-intensive, and development progress remains slower than anticipated. These delays have raised concerns from program officials about the prime contractor’s ability to complete the program’s software in a timely manner. Software risks remain because, notwithstanding how long the program has been in development, the Air Force and Sentinel contractor have yet to finalize software design or software development metrics, and are replanning the delivery schedule. Benefits of the new system. Sentinel program officials expect that Sentinel will provide the nation with a significantly more capable ICBM system with modular capacity to adapt as threats and technology evolve. Further, officials stated that the Air Force and the Office of the Secretary of Defense are actively mitigating risks to ensure there are no capability gaps during the Minuteman III to Sentinel transition. Our key findings from our 2025 report on the Air Force’s planning for the transition to Sentinel while concurrently operating Minuteman III include the following: Opportunity to improve risk management. The transition from Minuteman III to Sentinel involves a complex, total weapon system replacement. But the Air Force hasn’t developed a risk management plan for the most complex project the service has ever undertaken. A very large project that costs $1 billion or more, affects 1 million or more people, and runs for years may be referred to as a megaproject. Megaprojects are extremely risky ventures, notoriously difficult to manage, and often fail to achieve their original objectives. Research has found that planners can better position a megaproject for success by investing appropriate time at the front end to develop the tools and processes to identify and manage risks. Components of the Minuteman III Intercontinental Ballistic Missile System Sustainment risks for Minuteman III. As a result of delays to Sentinel, the Air Force may need to operate Minuteman III through 2050, 14 years longer than planned. Prolonged operation of the aging system presents sustainment risks. Addressing these risks in a transition risk management plan would help ensure the system meets requirements during the transition. Flight testing risks for Minuteman III. Sentinel delays also mean the Air Force will need to carefully manage the supply of parts for Minuteman III flight testing. The Minuteman III program conducts multiple flight tests per year to provide confidence in weapon system performance. The Air Force will need sufficient time to ensure the necessary parts are available for prolonged testing. A coordinated ICBM flight test plan would enable the Air Force to determine the quantity and type of parts the service needs to be able to test Minuteman III for another 2 decades. Training for Air Force security forces. DOD will need to complete Sentinel launch facility test and evaluation activities early in the transition to inform DOD and Air Force security policy updates. Because security forces incorporate these updates into unit-level operating instructions, these policy updates will be needed to train Air Force security forces for the transition. The Big Picture The Air Force aims to replace the 50-year-old Minuteman III intercontinental ballistic missile (ICBM) system, part of the nation’s strategic nuclear deterrent, with the new and upgraded Sentinel system. But Sentinel is years behind schedule and proving far costlier than expected. In 2024, the Air Force notified Congress that the Sentinel program had breached a statutory cost growth threshold—known as a Nunn-McCurdy unit cost breach—which triggered a statutorily required review process. The Nunn-McCurdy statute is a tool for Congress to use to hold the Defense Department (DOD) accountable for unit cost growth on major defense acquisition programs. The Air Force is restructuring the Sentinel program to try and improve affordability. The Air Force estimated that the program would cost at least $141 billion, but the actual costs remain uncertain. Sentinel delays mean the Air Force will need to operate the aging Minuteman III longer than planned. The Minuteman III system consists of more than 600 facilities across five states including missile silos and command bunkers, all of which must be replaced. Air Force leadership has described this megaproject as the most significant and complex infrastructure project in the service’s history. Sound planning will be essential for the Air Force to successfully manage a project of this scale while meeting round-the-clock nuclear deterrent requirements. Challenges and Opportunities DOD and the Air Force are working to restructure the Sentinel program to ensure the program meets cost, schedule, and performance requirements. While working to obtain new Milestone B approval, the Air Force has the opportunity to correct program deficiencies that led to the 2024 Nunn-McCurdy unit costs breach, including by developing an integrated master schedule, maturing software development tools and processes, and finalizing the design of the Sentinel launch facility. The Air Force can also use the Sentinel program’s restructuring period to bolster its project management framework, as we recommended, to better position the service for a successful transition from Minuteman III to Sentinel. Key recommendations from our 2025 report on the transition: ➢ The Air Force should develop a transition risk management plan, including a plan to address Minuteman III sustainment risks. ➢ The Air Force should expeditiously develop a post-2030 Minuteman III operational test launch plan that is aligned with a Sentinel fielding plan. ➢ The Air Force should develop a plan to complete the necessary Sentinel launch facility test and evaluation activities that will inform DOD security policy updates. For more information, contact Joseph W. Kirschbaum at KirschbaumJ@gao.gov.