AI in federal financial management Treasury IRS and OMB actions

What is happening

• Executive Order 14110 established government-wide direction for safe, secure, and trustworthy AI, tasking OMB to set implementation guidance for agencies and requiring public AI use case inventories to increase transparency of AI use across missions including financial management and tax administration¹.
• OMB M-24-10 operationalizes EO 14110, directing agencies to designate Chief AI Officers, establish AI Governance Boards, categorize and manage risks of AI use (including rights-impacting functions), and publish annual AI inventories; the memo aligns agency AI risk practices to the NIST AI RMF 1.0²³⁴.
• IRS’s Strategic Operating Plan commits to expanding the use of advanced analytics and machine learning to improve compliance (including complex business return examinations) and taxpayer services, establishing a concrete pathway for AI-enabled capabilities in tax administration⁵.
• Treasury’s Bureau of the Fiscal Service operates the Do Not Pay program to support payment integrity across agencies with data-driven analytics and risk checks, providing an enterprise capability relevant to improper payment prevention in federal financial management⁶.

Why it matters for federal financial management

• OMB’s governance requirements apply to AI that influences benefits, payments, eligibility, or enforcement outcomes, which are core to federal financial management; agencies must inventory such AI uses, assess impact (including rights-impacting), and implement safeguards consistent with NIST’s RMF functions (Govern, Map, Measure, Manage)²³.
• IRS’s planned deployment of analytics and machine learning to detect noncompliance affects revenue collection accuracy and fairness, while AI-informed taxpayer service can reduce burden and improve responsiveness; both must meet OMB’s transparency, evaluation, and human oversight expectations²⁵.
• Payment integrity initiatives that incorporate algorithmic risk assessment must align with PIIA requirements and OMB’s guidance for preventing and reporting improper payments, integrating AI risk controls with financial management internal control frameworks²⁷.

Agency-by-agency actions

Office of Management and Budget

• Governance: M-24-10 mandates Chief AI Officers, AI Governance Boards, and annual AI inventories; it requires documentation of intended use, benefits, and risks, and establishes requirements for AI that materially impacts the public’s rights or safety (e.g., identity verification, eligibility determinations, enforcement prioritization) including testing, monitoring, and human oversight².
• Standards alignment: M-24-10 directs agencies to leverage NIST’s AI RMF 1.0 for risk management, establishing a common vocabulary and practices for trustworthy AI across the federal enterprise²³.
• Transparency: OMB coordinates publication of agency AI inventories via AI.gov, making use cases visible to the public and enabling cross-agency learning and scrutiny⁴.

Treasury Department (including Bureau of the Fiscal Service)

• Payment integrity operations: Treasury’s Do Not Pay program provides agencies with analytic tools and data checks to help prevent and detect improper payments before disbursement, supporting CFO Act agencies’ compliance with PIIA and OMB A-123/PIIA guidance; agencies integrating AI into these workflows must treat models and risk scores as part of internal controls under OMB’s AI governance⁶⁷².
• Department-wide AI inventories: Treasury participates in the OMB-required AI inventory process, which surfaces AI-enabled use cases across Treasury bureaus; inventories are published through AI.gov⁴.

Internal Revenue Service

• AI-enabled compliance: IRS’s Strategic Operating Plan commits to employing analytics and machine learning to improve case selection and compliance in complex areas (e.g., partnerships and large corporations), and to enhance detection of noncompliance; IRS ties these capabilities to Inflation Reduction Act investments in modernizing data and systems⁵.
• AI in taxpayer services: The plan also targets expanded digital services, including intelligent automation and data-driven assistance, which must comply with OMB M-24-10 requirements for transparency, evaluation, and appropriate human oversight when AI affects taxpayer interactions⁵².

Implementation implications for missions

• Inventory discipline: Financial management leaders should ensure all AI or ML that influences payments, eligibility, collections, enforcement prioritization, or taxpayer service is captured in the agency’s AI inventory with clear descriptions of function, data, human oversight, and evaluation practices; AI.gov publication creates external accountability²⁴.
• Risk controls: Apply NIST AI RMF practices—define context and harms, establish measurable performance and robustness metrics, and implement monitoring and incident management—alongside financial internal controls under OMB’s PIIA guidance; treat AI model governance as part of A-123 control evaluations³⁷.
• Rights-impacting guardrails: For AI used in identity proofing, eligibility determinations, or enforcement triage, implement pre-deployment testing, bias and error analysis, appeal or override pathways, and human-in-the-loop decisioning consistent with M-24-10 requirements².
• Cloud compliance: Host AI workloads involving PII and financial data in FedRAMP-authorized environments; Azure Government holds a FedRAMP High authorization and provides security and compliance controls relevant to OMB and NIST requirements, which agencies can enforce via Azure Policy and related governance tooling⁸⁹¹⁰.
• Model and service portfolio: Where generative or ML services are needed, Azure OpenAI Service availability in Azure Government provides agencies a path to deploy foundation models in compliant enclaves; responsible AI documentation should be mapped to NIST RMF artifacts to demonstrate governance and risk mitigation¹¹¹²³.
• Continuous oversight: As IRS scales AI-enabled analytics for enforcement and service, program owners should implement ongoing evaluation, drift monitoring, and post-deployment incident response, publishing updates in inventories and governance board minutes consistent with M-24-10²⁵⁴.

Decision points for federal leaders

• Scope AI inventories to include algorithmic triage and risk scoring used in payment integrity and enforcement—even when embedded in analytics workflows—to meet OMB transparency expectations²⁴.
• Require NIST RMF-aligned artifacts (risk registers, evaluation protocols, monitoring plans) as part of A-123 control documentation for AI-enabled financial processes³⁷.
• Select cloud enclaves with FedRAMP High and mission accreditations for AI workloads handling PII and financial data; enforce policy guardrails with platform-native controls and document them in governance board oversight⁸⁹¹⁰.

¹: Executive Order 14110 — Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence — https://www.whitehouse.gov/briefing-room/presidential-actions/2023/10/30/executive-order-on-the-safe-secure-and-trustworthy-development-and-use-of-artificial-intelligence/ ²: OMB Memorandum M-24-10 — Advancing Governance, Innovation, and Risk Management for Agency Use of Artificial Intelligence — https://www.whitehouse.gov/omb/memoranda/2024/m-24-10/ ³: NIST Artificial Intelligence Risk Management Framework (AI RMF 1.0) — https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.1270.pdf ⁵: IRS Strategic Operating Plan — Transforming Tax Administration — https://www.irs.gov/strategic-operating-plan ⁴: AI.gov — Agency AI Use Case Inventories — https://ai.gov/inventories/ ⁶: Treasury Bureau of the Fiscal Service — Do Not Pay Program — https://www.donotpay.treasury.gov/ ⁷: OMB Memorandum M-21-19 — Compliance with the Payment Integrity Information Act of 2019 — https://www.whitehouse.gov/wp-content/uploads/2021/03/M-21-19.pdf ⁸: FedRAMP Marketplace — Microsoft Azure Government — https://marketplace.fedramp.gov/#!/product/microsoft-azure-government ⁹: Azure Government security and compliance overview — https://learn.microsoft.com/en-us/azure/azure-government/compliance/azure-government-security-compliance ¹¹: Azure OpenAI Service expands to Azure Government — https://azure.microsoft.com/en-us/blog/azure-openai-service-expands-to-azure-government/ ¹²: Microsoft Responsible AI documentation — https://learn.microsoft.com/en-us/azure/ai/responsible-ai/overview ¹⁰: Azure Policy overview — https://learn.microsoft.com/en-us/azure/governance/policy/overview