Based on recent developments in federal AI strategy, several U.S. federal agencies are planning to use AI data centers and related technologies to ensure the accuracy of their data. A major driver of these efforts is the White House's "America's AI Action Plan," which was outlined in executive orders and policy memos released in 2025.
Government-wide initiatives
New data center infrastructure: As part of the AI Action Plan, the Department of Energy (DOE) has selected federal sites—including the Idaho National Laboratory and Oak Ridge Reservation—to host new AI-focused data centers. These centers will provide the massive computing power necessary for advanced AI analytics.
Government-wide AI suite: The General Services Administration (GSA) launched USAi.gov, a secure generative AI evaluation suite for federal agencies. The platform allows agencies to test and adopt AI tools while complying with federal standards, with GSA emphasizing that AI depends on "high-quality, well-governed data".
Procurement guidance: The Office of Management and Budget (OMB) has issued new guidance for agencies on how to procure AI systems, including measures to ensure data accuracy. This includes requiring vendors to provide documentation that facilitates transparency and accountability.
Addressing cybersecurity risks: The Government Accountability Office (GAO) has highlighted the significant cybersecurity risks associated with AI, including data poisoning. The White House's plan addresses this by mandating new technical standards for high-security AI data centers, which will be led by the Department of Homeland Security (DHS) and the National Institute of Standards and Technology (NIST).
Specific agency plans for data accuracy
General Services Administration (GSA): The GSA's AI compliance plan ensures data accuracy through a Chief AI Officer (CAIO) who maintains a comprehensive and updated inventory of all AI use cases. The agency emphasizes that trustworthy AI depends on accurate, traceable, and reusable data.
Internal Revenue Service (IRS): The IRS is exploring AI to improve tax compliance and identify inconsistencies in tax filings by analyzing large datasets.
Department of Defense (DoD): The DoD uses AI for applications like predictive maintenance and military logistics. Its strategy emphasizes expanding its AI capabilities with careful governance and security measures.
Securities and Exchange Commission (SEC) and Federal Trade Commission (FTC): These agencies use AI to detect fraud and financial misconduct by analyzing large datasets in real-time.
Department of Energy (DOE) and National Nuclear Security Administration (NNSA): Following the selection of federal lands for AI data centers, both agencies have issued requests for proposals from private companies. A key goal of these partnerships is to build next-generation data centers to advance AI and energy infrastructure.
Challenges
to ensuring data accuracy
Despite these efforts, federal agencies face significant challenges in managing AI, according to a July 2025 GAO report. These include:
Incomplete inventories: A 2023 GAO report found that most agencies' AI use-case inventories were not fully comprehensive or accurate.
Lagging
policies: Agencies have struggled to keep their internal policies updated to
match the rapid pace of AI development.
Insufficient resources: Many agencies lack sufficient funding and technical talent to fully address data accuracy and AI deployment challenges.
Data contamination risks: Cybersecurity experts warn that risks like data poisoning threaten the reliability of AI models.
In recent efforts to expand AI use, US federal agencies plan to use AI data centers to ensure data accuracy in fields like finance, national security, and public health. The effort is part of a broader federal strategy, outlined in the July 2025 "America's AI Action Plan" and related executive orders, to establish new, high-security AI data centers on federal lands. These data centers will be integral to developing more reliable AI systems by enabling control over data integrity and the use of technical standards.
Agencies
involved in developing AI data centers
Several federal departments are involved in accelerating the creation of AI data centers for developing and implementing secure AI systems.
Department of Defense (DoD): The DoD is involved in creating technical standards for high-security AI data centers and is exploring using military bases for commercial AI data center development. The DoD's Chief Digital and Artificial Intelligence Office (CDAO) is responsible for accelerating the adoption of AI capabilities across the department.
Department of Energy (DOE): The DOE has announced the selection of four federal sites for new AI data center and energy co-location projects, with initial proposals due in November 2025. The selected sites include the Idaho National Laboratory, Oak Ridge Reservation, Paducah Gaseous Diffusion Plant, and the Savannah River Site.
Department of Commerce (DOC): The DOC's National Institute of Standards and Technology (NIST) is leading the development of AI standards and evaluation metrics across various sectors.
Department of Homeland Security (DHS): The DHS is involved in promoting the sharing of AI security threat information and intelligence. It also leverages AI for various functions, including enhancing threat detection, analyzing cyber events, and verifying proof of life in applications like the CBP One app.
Agencies
using AI for data accuracy
Beyond the development of AI infrastructure, numerous agencies are incorporating AI into their daily operations to improve data accuracy.
Internal Revenue Service (IRS): The IRS uses AI to enhance tax compliance by identifying inconsistencies and potential fraud in tax filings.
Securities and Exchange Commission (SEC): The SEC uses AI to monitor capital markets and analyze large datasets to detect fraud and other forms of financial misconduct.
Centers for Disease Control and Prevention (CDC): The CDC uses AI for public health applications, such as using natural language processing to analyze text-based information for outbreak detection and using machine learning to track health trends.
Department of Health and Human Services (HHS): HHS employs AI-enabled tools to combat fraud, waste, and abuse in programs like Medicare and Medicaid.
General
Services Administration (GSA): The GSA launched USAi in August 2025, a
platform that provides federal agencies with secure access to AI evaluation
tools for testing and adopting the technology safely and efficiently.
Government-wide initiatives
Office of Management and Budget (OMB): The OMB issues government-wide guidance for AI use, including mandates for agencies to produce AI use case inventories.
Government Accountability Office (GAO): The GAO has developed an AI accountability framework that includes key practices for ensuring data quality and reliability when using AI. Recent GAO reports, however, have highlighted that many agency AI inventories are incomplete or inaccurate.
The "America's AI Action Plan": This plan treats security and data integrity as fundamental to AI success. It mandates the creation of technical standards for AI data centers and requires agencies to use systems that can prove data provenance and maintain audit trails to ensure the trustworthiness of AI systems.
Norb Leahy, Dunwoody GA Tea Party Leader
No comments:
Post a Comment