AI Governance Institute logo
AI Governance Institute

Practical Governance for Enterprise AI

· CMP-008Medium effort

Federal AI Regulatory Monitoring and Pre-Deployment Vetting

Monitor US federal AI regulatory developments across executive orders, agency guidance, and frontier model requirements, and maintain a pre-deployment vetting protocol aligned to current federal expectations.

Objective

Ensure AI systems deployed in US federal contexts, or by organizations subject to US federal AI requirements, are vetted against current federal standards before deployment and that the organization receives timely notice of new federal AI obligations.

Maturity Levels

1

Initial

Federal AI regulatory developments are monitored ad hoc with no formal tracking or vetting process.

2

Developing

Key executive orders are tracked, but agency-level AI guidance and sector-specific rulemaking are not systematically monitored.

3

Defined

A federal regulatory monitoring process covers executive orders, NIST AI program outputs, FTC guidance, sector agency rulemaking, and OMB memoranda. A pre-deployment checklist is reviewed before any AI system is deployed in a federal or regulated context.

4

Managed

Federal regulatory monitoring feeds into the unified compliance register. The pre-deployment checklist is updated within 30 days of material federal guidance changes. A federal compliance lead owns the process.

5

Optimizing

The organization participates in NIST AI program stakeholder processes and federal agency AI working groups. Pre-deployment vetting is automated for standard deployment types.

Evidence Requirements

What an auditor or assessor would expect to see for this control.

  • Federal regulatory monitoring log covering the past 12 months, with each material federal AI development documented and assessed.
  • Pre-deployment vetting checklist completed for each AI system deployed in a federal or regulated context, with sign-off date.

Implementation Notes

Key steps

  • Establish a federal regulatory monitoring scope covering:

    • Executive orders (EO 14110 rescinded; EO 14179 on AI leadership; subsequent orders)
    • OMB memoranda on AI use in federal agencies (M-24-10, M-24-18 successors)
    • NIST AI RMF updates, profiles, and playbooks
    • FTC enforcement guidance on AI representations and automated decision-making
    • Sector-specific agency AI rulemaking: FDA (AI/ML software as medical device), OCC and Federal Reserve (model risk management), SEC (AI disclosure), FAA (autonomous systems)
    • Federal Acquisition Regulation AI clauses for government contractors
  • Subscribe to monitoring sources: Federal Register AI-tagged notices, NIST AI newsletter, OMB policy announcements, FTC blog and enforcement actions.

  • Define a pre-deployment vetting checklist for AI systems entering federal or regulated contexts:

    • Does the system comply with current NIST AI RMF requirements for the applicable use case?
    • Has a bias and fairness assessment been completed?
    • Is the system documented to the applicable federal technical documentation standard?
    • Have export control and procurement rules been reviewed if the system uses foreign technology?
    • Is a human oversight mechanism in place meeting federal expectations for the use case?
  • Update the checklist within 30 days of any material federal guidance change.

Key current federal requirements

  • NIST AI RMF is the de facto federal standard and is cited in OMB guidance for agency AI use.
  • EO 14179 (Jan 2025) established US AI leadership as a priority; removed some earlier safety requirements but maintained others through agency guidance.
  • FTC treats material misrepresentations about AI capabilities as deceptive trade practices under Section 5.

Example Implementation

Federal AI Pre-Deployment Vetting Checklist

System: [System name] Version: [x.x] Deployment context: [Federal agency / regulated sector] Completed by: [Name] Date: [Date]

RequirementStandardStatusEvidenceNotes
NIST AI RMF GOVERN function documentedNIST AI RMF 1.0PassAI governance policy v3.1
NIST AI RMF MAP function: risk identification completeNIST AI RMF 1.0PassRisk register entry 2026-04
Bias and fairness assessment completedOMB M-24-10PassFairness report 2026-03
Technical documentation meets federal standardsNIST AI 100-1In progressDraft docNeed sign-off
Human oversight mechanism documentedEO 14110 successors / agency guidancePassHITL policy v2
Export control review completedEAR / ITARN/ASystem uses no controlled tech
FTC capability representation reviewFTC Section 5PassMarketing review 2026-02

Control Details

Control ID
CMP-008
Domain
Typical owner
Legal / Compliance / Government Affairs
Implementation effort
Medium effort
Agent-relevant
No

Tags

federal AI regulationexecutive orderspre-deployment vettingNIST AI RMFFTCUS AI policy

Related Playbook

How do we document AI decision-making for auditability?How do we disclose AI governance maturity to investors and regulators?Who owns AI governance within the organization?How do we build an AI governance program from scratch?What do we do when an AI system causes harm or fails?How do we govern AI models from preview release through retirement?Is our AI red-teaming rigorous enough?How do we govern our AI supply chain and manage upstream model dependencies?How do we detect and mitigate algorithmic bias?What does audit-ready AI documentation look like in practice?How do we report AI risk to the board and audit committee?How do we comply with China's AI regulations?How does the EU AI Act affect our global operations?How do we govern AI agents that take autonomous actions?How do we comply with the EU AI Act?How do we perform an AI risk assessment?How are we managing third-party AI risks?How do we manage third-party AI vendors safely throughout the vendor lifecycle?How do we measure and mitigate algorithmic bias?What is our process for model drift monitoring?How do we build and maintain a multi-framework AI risk register?How do we map AI compliance obligations across multiple jurisdictions?How do we engage regulators and standards bodies proactively on AI governance?What are our obligations under emerging AI regulations?How do we ensure third-party AI vendors meet our standards?How do we monitor voluntary AI safety commitments and respond when they change?