Governance & Program Info
Wargaming & Simulation AI — Governance
Cross-builder institution context and per-item ownership, due dates, status, and next actions for the governance-relevant checklist items in this builder.
Institution context
Program info
Applies across every builder in the app. Stored locally; nothing leaves the browser.
Checklist governance
Items (0 of 14 marked complete)
Annotate ownership, due date, status, and next action. Items on the left come from the builder's governance / compliance phases.
05 · Governance & Compliance
Classification guide for trained red-force models
A model trained on classified TTPs inherits that classification. Document the guide that will travel with every weight file, checkpoint, and inference artifact.
05 · Governance & Compliance
Enclave and cross-domain handling for training and inference
Training on classified data, serving to a classified exercise — confirm the end-to-end enclave posture and any cross-domain guards.
05 · Governance & Compliance
Data-spill and incident response plan for training data
What happens if unclassified training data is discovered to contain classified material, or if a classified weight leaks to a lower enclave.
05 · Governance & Compliance
Coalition / FVEY / NATO releasability review
If the AI will ever train alongside coalition partners, a releasable analog must be defined now — retrofitting releasability is expensive and error-prone.
05 · Governance & Compliance
ITAR USML categorization of the AI capability
Determine and document the USML category the AI capability falls under — and any EAR 600-series considerations.
05 · Governance & Compliance
DoDD 3000.09 review where red-force autonomy tools are transferable
Where red-force autonomy logic could transfer into a weapon system, DoDD 3000.09 (autonomy in weapon systems) review is appropriate even if the AI itself is a simulation artifact.
05 · Governance & Compliance
LOAC review of simulated conduct
Legal / JAG review that scenarios and AI behaviors do not inadvertently train the audience to violate the Law of Armed Conflict.
05 · Governance & Compliance
Alignment to DoD AI Ethical Principles
Responsible, Equitable, Traceable, Reliable, Governable — documented alignment to the five DoD AI Ethical Principles adopted in 2020.
05 · Governance & Compliance
Traceability and decision provenance
Every AI decision logged and traceable back to model version, seed, scenario state, and feature inputs — required for both VV&A and post-event legal review.
05 · Governance & Compliance
Governable / human-on-the-loop controls
Exercise controllers must be able to pause, override, or constrain AI behavior during an event — the AI is never the final authority inside a training environment.
05 · Governance & Compliance
DoDI 5000.82 digital capability acquisition alignment
Map the AI capability to DoDI 5000.82 requirements for acquisition of digital capabilities — particularly where the AI becomes a program-of-record sustainment item.
05 · Governance & Compliance
DoDI 5000.83 T&E-for-software alignment
T&E Master Plan that treats AI behavior as a testable, monitorable characteristic — not a one-time acceptance event.
05 · Governance & Compliance
Change management for retrained / updated models
Who approves a retrain, what regression suite must pass, how is the old version retained for replay, what is the rollback procedure.
05 · Governance & Compliance
Sustainment funding model for continuous red-force improvement
A stale red force becomes predictable and loses training value. Budget for ongoing retraining as new intelligence arrives is a governance decision, not a technical one.