U.S. Department of the Treasury

On March 23, 2025, President Trump signed Executive Order (EO) 14303, “Restoring Gold Standard Science.” The order seeks to “restore the American people’s faith in the scientific enterprise and institutions that create and apply scientific knowledge in service of the public good.”

The order tasked federal agencies to report to the White House’s Office of Science and Technology Policy (OSTP) on how they would address each tenet of Gold Standard Science (GSS) as laid out in EO 14303 for the conduct and management of their respective scientific activities. Treasury has been committed to, and reaffirms its commitment to, applying gold standard principles to its functions as rigorous, transparent, and impartial analysis this build public confidence in policy decisions that have real impacts on the lives of Americans.

Introduction

The mission of the U.S. Department of the Treasury (Treasury) is to maintain a strong economy, to create economic and job opportunities through economic growth and stability at home and abroad, to strengthen national security, and to manage the U.S. Government’s finances and resources effectively. In short, Treasury’s role is to be a good steward of the U.S. economy and financial system and to promote the United States’s influence in the world economy.

On March 23, 2025, President Trump signed Executive Order (EO) 14303, “Restoring Gold Standard Science.”¹ The order seeks to “restore the American people’s faith in the scientific enterprise and institutions that create and apply scientific knowledge in service of the public good.” The order tasked federal agencies to report to the White House’s Office of Science and Technology Policy (OSTP) on how they would address each tenet of Gold Standard Science (GSS) as laid out in EO 14303 for the conduct and management of their respective scientific activities. It is imperative that agencies produce accurate and reliable data and that they exercise rigor, transparency, and impartiality in consequent analyses. This will build public confidence in policy decisions that have real impacts on the lives of Americans.

Although Treasury’s mission does not directly relate to applied research or experimental development (as defined in OMB Circular No. A-11, it uses basic research to fulfill its role as steward of the U.S. economy and financial systems. Treasury has been committed to, and reaffirms its commitment to, applying gold standard principles to its functions. The purpose of this report is five parts:

Metrics and Mechanisms of GSS Implementation

Treasury is still exploring the methods by which it will evaluate GSS implementation. Research and research-adjacent offices will collaborate the support and evaluation offices, like the Office of Strategic Planning and Performance Improvement, to ensure implementation of GSS principals and report to the OSTP in the annual reports due on September 1 each year starting in 2026.

Emerging Technology Application

Artificial Intelligence (AI) has the potential to transform society, and Treasury is in the process of embracing the potential gains from AI to better serve the American people. Treasury’s AI Strategy reflects both the department’s mission and the moment by applying AI to modernize Treasury operations, strengthen American economic security, and reinforce the Department’s global leadership in financial innovation and governance.

Treasury’s AI Strategy is anchored in three core objectives:

• Apply AI to improve Treasury operations, deploying AI to enhance analytic capabilities, to improve service delivery, and to increase the efficiency of Treasury programs — while safeguarding classified and sensitive information.
• Advance responsible AI adoption in the financial sector by working with financial institutions, technology partners, and regulators to support innovation while safeguarding market integrity and operational resilience.
• Promote U.S. leadership through global and federal AI alignment. Treasury will engage with interagency and international partners to shape the standards and frameworks that govern AI adoption across financial systems around the world.

Treasury’s AI Governance Board is working with career and appointees to explore how AI implementation across the department can improve consistency, coordination, and accountability of Treasury operations and analyses used for policymaking.

Challenges

Economic, tax, and financial analyses and research are often not experimental. Rather they routinely use existing data and apply accepted statistical techniques to ascertain correlation and causality, magnitude, uncertainty, etcetera. This can lead to several challenges in applying GSS tenets.

• The biggest challenge for applying GSS at Treasury is that the department works in a quickly changing economic and financial environment that may require prompt action. The usefulness of analysis to inform public policy is constrained by the necessity of timeliness. At the same time, accuracy and validity of the models used in analysis must not be sacrificed for timeliness, but the application of some GSS tenets is outweighed by the needs of the moment.
• Another challenge for applying GSS at Treasury lies in the very nature of data used for policy analysis. For most economic and financial research, Treasury does not collect its own data, but rather relies on data from various sources, such as surveys conducted by the Census Bureau, Bureau of Labor Statistics, Federal Reserve Board and regional banks, etcetera as well as private sector data. The results of analysis can be limited by unknown bias within the data, which may not be apparent a priori and for which analysts may not sufficiently control. As a result, conclusions can change based on (sometimes large) data revisions.
• Data availability is another key limitation to the pursuit of longer form social policy research at Treasury. While aggregate survey data are publicly available from federal statistical agencies, data are not usually disaggregated finely enough to provide confidence in results—thus limiting the implementation of some GSS tenets. Researchers can apply to use confidential microdata from Federal Statistical System Research Data Centers (RDC), the cost is prohibitively expensive—especially in a budget constrained environment. Increased access by qualified Treasury staff to RDCs would provide for better analysis of discrete policy questions for principals, and it would provide Treasury increased abilit
• Data that are more disaggregated are often released on a lagged timeline. For example, the latest Statistics of Income (SOI) data were released in 2025 and pertained to tax returns in the 2022 tax year. This adds uncertainty as to how applicable analysis using the 2022 SOI data are to current policy. Replicability, however, is not affected if analysis is repeated annually. Repeated analysis with new data can confirm or cast doubt on previous estimates and related error bands, updating the knowledge research provides for principals in pursuing public policies.

¹ Federal Register, Vol. 90, No. 102. Executive Order 14303 of May 23, 2025. “Restoring Gold Standard Science.” Available: https://www.govinfo.gov/content/pkg/FR-2025-05-29/pdf/2025-09802.pdf