AI Assurance Directory

Path

I'm a DoD contractor using AI with CUI

Cited starting points when a DoD contractor’s AI system processes, stores or transmits controlled unclassified information: CMMC, NIST SP 800-171, and DoD Responsible AI. Sources last checked 15 October 2026.

Cited intro

CMMC and NIST SP 800-171 apply to information systems, not to technologies by name. An AI system is in scope when it processes, stores or transmits controlled unclassified information (CUI) or protects such a system. We searched 32 CFR Part 170 and found no AI-specific CMMC clause.

DFARS 252.204-7021 requires a current CMMC status at the level the contracting officer specifies for systems used in performance that process, store or transmit FCI or CUI. DoD CIO says Phase II was suspended on 13 July 2026 and that Phase I self-assessment requirements remain in place. The same page also says implementation is paused in Phase 1. Buyer-intent write-ups: CMMC and AI tools handling CUI, DoD 3000.09 checklist for autonomy vendors, ISO 42001 vs NIST AI RMF for defense contractors, and choosing AI red-teaming tools.

Frameworks

  • NIST AI RMF

    A voluntary, sector-agnostic US framework that organizes AI risk work into four functions - Govern, Map, Measure and Manage.

  • CMMC / NIST SP 800-171

    How DoD contractor cybersecurity rules apply to AI systems that process controlled unclassified information. The rules are technology-neutral, and CMMC Phase 2 is currently suspended.

  • DoD Responsible AI

    The Department of Defense's five AI ethical principles, its Responsible AI Strategy and Implementation Pathway, and the public CDAO toolkit (released as RAI, now titled AIA Toolkit).

  • DoD Directive 3000.09

    DoD policy for autonomous and semi-autonomous weapon systems. It requires appropriate human judgment over the use of force, V&V and T&E, and senior review. It does not cover unarmed or non-weapon AI.

  • DoD RAI Strategy

    The Department’s RAI Strategy and Implementation Pathway operationalizes the five AI ethical principles. The June 2022 pathway names 64 lines of effort. It is Department direction, not a product certificate.

  • CDAO RAI Toolkit

    CDAO’s public self-evaluation toolkit for aligning AI projects to DoD (now DoW) AI Ethical Principles. Released November 2023 as the RAI Toolkit; the live Tradewinds page titles it the AIA Toolkit.

  • NIST SP 800-218A

    A July 2024 SSDF Community Profile that adds generative-AI and dual-use foundation-model practices to NIST SP 800-218. Guidance for model producers, system producers and acquirers — not a certification.

Articles

Listed tools

  • CDAO AI Assurance (AIA) Toolkit

    DoD Chief Digital and Artificial Intelligence Office · Government-published · Government-published resource

    CDAO self-evaluation toolkit that operationalizes DoW AI Ethical Principles. Released November 2023 as the RAI Toolkit; the live Tradewinds page titles it AIA Toolkit. For Program Managers, AI Developers, Acquisition Teams, Legal/Compliance and End Users; IL2 and IL5.

  • CDAO Joint AI Test Infrastructure Capability (JATIC)

    DoD Chief Digital and Artificial Intelligence Office · Government-published · Government-published resource

    CDAO program (funded FY23–FY29) of Python AI T&E libraries, the RAVEN MLOps platform and the maite types/protocols. Focused on CV classification and object detection for UAV, satellite and autonomous-driving missions. Public packages on GitHub, Conda and PyPI. Not a certification.

  • CDAO Test and Evaluation of Artificial Intelligence Models Framework

    DoD Chief Digital and Artificial Intelligence Office · Government-published · Government-published resource

    April 2024 public CDAO primer on what to consider in a T&E strategy for AI models in a Defense context. Four focus areas, a performance iceberg beyond correctness, and named links to the DoD Ethical AI Principles and NIST AI RMF. Explicitly not binding policy.

  • Counterfit

    Microsoft · Open source · AI assurance & verification · Open source · MIT

    MIT-licensed CLI (banner Version 1.1.0) that wraps Adversarial Robustness Toolbox, TextAttack and AugLy to assess ML-system security. Linux or Windows WSL; optional Azure Container Instance. Python 3.8+. Windows and macOS native support is still listed as work in progress.

  • Giskard

    Giskard · Commercial · AI assurance & verification

    LLM-agent testing (Python SDK and Hub UI), red-teaming and runtime Guards with an EU AI Act Policy Pack (OPA/Rego). On-premise Hub for public-sector and defense workloads. Named sectors: banking and insurance.

  • HEART (Hardened Extension of ART)

    IBM (CDAO JATIC collaboration) · Open source · AI assurance & verification · Open source · MIT

    MIT-licensed Python 3.9–3.11 library (pip install heart-library; 0.7.0) that extends ART for DoD T&E. Modules: attacks, estimators, metrics. Aligns to MAITE; depends on ART ≥1.20.1. CDAO contract W519TC-23-9-2037. Optional Gradio front end.

  • HiddenLayer AI Security Platform

    HiddenLayer · Commercial · AI assurance & verification

    AI security platform with government and defense pages: NIST AI RMF alignment, contractor and open-source model scanning, real-time monitoring, and air-gapped options for classified disconnected environments. The homepage also names AI Discovery, supply-chain, attack simulation and runtime security.

  • NIST AI RMF Playbook

    NIST AI Resource Center · Government-published · Government-published resource

    NIST AI RMF Playbook of suggested actions aligned to Govern, Map, Measure and Manage sub-categories in AI RMF 1.0. Neither a checklist nor a set of steps. Suggestions are voluntary. Downloadable as PDF, CSV, Excel and JSON. The Playbook will be updated after AI RMF 1.0 is revised.

  • NIST ARIA (Assessing Risks and Impacts of AI)

    NIST Information Technology Laboratory · Government-published · Government-published resource

    NIST ITL evaluation (Assessing Risks and Impacts of AI): sector- and task-agnostic model testing, red-teaming and field testing. LLM pilot. Schedule Kickoff May 2024 through Summer 2025 Pilot Summary Report Out. That report is still listed as coming soon.

  • Promptfoo

    Promptfoo · Open source · AI assurance & verification · Open source · MIT

    Open-source gen-AI red-teaming tool that scans 50+ vulnerability types and maps plugins to the NIST AI RMF, ISO/IEC 42001 and EU AI Act. Runs locally against HTTP APIs, browsers or models. MIT-licensed.

  • PyRIT

    Microsoft AI Red Team · Open source · AI assurance & verification · Open source · MIT

    Microsoft AI Red Team framework for automated and human-led generative-AI red teaming. Python 3.13 recommended; pip install pyrit. Ships Crescendo, TAP and Skeleton Key, CoPyRIT GUI, pyrit_scan/pyrit_shell, and SQLite or Azure SQL memory. MIT-licensed.

  • Resaro Approved Intelligence Platform

    Resaro · Commercial · AI assurance & verification

    On-premise, air-gapped TEVV platform that produces ASQI-structured deployment evidence for civil and defence AI. CI/CD-native workflows over text and image; used in the IMDA/AI Verify Global AI Assurance Pilot.

Also on this site

Questions

Does CMMC have an AI-specific level?

No source we cite says that. CMMC and NIST SP 800-171 apply when a system processes, stores or transmits CUI or FCI. The rules are technology-neutral.

Does DoD Directive 3000.09 apply because we use AI on a contract?

No. 3000.09 does not apply to autonomous or semi-autonomous systems that are not weapon systems. CUI handling still sits under CMMC and DFARS.

Is there a DFARS clause titled Artificial Intelligence or Super Intelligence that must flow down?

We searched live DFARS Part 252 text and did not find a clause titled that way. What primes actually flow down, on the pages we cite, are safeguarding clauses such as 252.204-7012, 252.204-7020, 252.204-7021 and FAR 52.204-21.

Sources for this intro

These official pages support the intro above. They are not endorsements.

Sources last checked on 15 October 2026.

  1. 32 CFR Part 170 — CMMC Program - eCFR
  2. 48 CFR 252.204-7021 (DFARS) - eCFR
  3. About CMMC - DoD CIO
  4. SP 800-171 Rev. 3 - NIST CSRC

Free next step: take the AI assurance readiness self-check or create a free account. Quoted evidence stays on Pro.