Governance & Compliance

AI Risk Management Framework — Govern, Map, Measure, Manage lifecycle functions

Implemented all 4 functions with 19 categories and 42 subcategories. Continuous monitoring active via automated telemetry.

Responsible, equitable, traceable, reliable, governable AI principles for defense systems

Full alignment with 5 ethical principles. Traceability logging covers all model inference paths. Human-in-the-loop for high-stakes decisions.

Adversarial Threat Landscape for AI Systems — tactics, techniques, and procedures

Covering 12 of 14 tactic categories. Active red-team benchmarks map to TTP matrix. Gaps in ML supply chain and physical-domain attacks.

High-risk AI system requirements — risk management, data governance, transparency, human oversight

Applicable for allied interoperability. Conformity assessment procedures partially implemented. CE marking pathway defined but not yet submitted.

AI Management System standard — Annex A controls for organizational AI governance

10 Annex A control categories mapped. Gap analysis completed. Certification audit targeted for Q3 2026. AIMS documentation 60% complete.

Executive Order on Safe, Secure, and Trustworthy AI — federal agency requirements

Dual-use foundation model reporting thresholds met. Red-team testing protocols implemented per NIST guidance. OMB M-24-10 compliance verified.

Insurance regulatory guidance on AI/ML usage — adopted by 24 states for governance of AI in insurance

Covers AIS Program requirements, third-party AI governance, bias testing, and consumer transparency. Maps to NIST AI RMF functions.

Federal Reserve / OCC model risk management guidance adapted for AI/ML systems

Model inventory, independent validation, ongoing monitoring, and outcome analysis. Extended to cover foundation model supply chain risks.