Regulation — federal / defense
DoD Directive 3000.09 — Autonomy in Weapon Systems
The 2023 update to DoD Directive 3000.09 extended the "appropriate levels of human judgment" language beyond kinetic weapon systems into the AI-inclusive-decision-support space. If your predictive-maintenance AI is on the timeline of a safety-of-flight incident, this directive is the one most likely to be cited.
What the directive actually says
"Autonomous and semi-autonomous weapon systems will be designed to allow commanders and operators to exercise appropriate levels of human judgment over the use of force."
DoD Directive 3000.09 §4.a — the core "appropriate levels of human judgment" mandate
"Persons who authorize the use of, direct the use of, or operate autonomous and semi-autonomous weapon systems must do so with appropriate care and in accordance with the law of war, applicable treaties, weapon system safety rules, and applicable rules of engagement."
DoD Directive 3000.09 §4.b — operator diligence obligation
What this means in plain English for predictive-maintenance AI
Predictive-maintenance AI is not itself a weapon system. But it produces decisions that gate mission-capable status — and mission-capable status feeds into use-of-force decisions. When the AI silently under-flags a mid-life combat asset as ready-for-mission, the "appropriate levels of human judgment" question moves from the AI-inclusive-decision-support layer into the operational layer:
- Did the AI produce a reliable readiness signal? If the sensor shows silent under-flagging, the answer for the drift-window is: no, not for the F3-shaped population.
- Did the human operator have the information to exercise appropriate judgment? If the aggregate fleet-readiness dashboard was green, the answer is: not without independent monitoring surfacing the differential.
The directive does not require the AI to be perfect. It requires the operator + program office to be able to demonstrate that appropriate human judgment was possible — which means the failure modes had to be knowable.
What triggers the exposure in the sample
F3 red_ground_urgent lane rate dropped from 34% to 8% silently over 45+ days. Approximately 270 F3 assessments that baseline expectations would have red-lane routed were down-scored to yellow-watch or green-ready. Any one of those, in the wrong operational scenario, becomes a JAG accident-investigation-board timeline entry. The directive's "appropriate levels of human judgment" test is applied against what was knowable and when.
The finding does not depend on any specific incident having occurred. The directive is about the design + operating stance, not only the outcome.
What the $499 Snapshot shows against this directive
- Per-asset-group distributional-shape analysis — the "was it knowable" evidence
- Dated first-drift-day + severity trajectory — the "when was it knowable" record
- 3 fix-first items including freeze / hold-for-flight-surgeon manual-review + Contracting Officer notification protocol — the "appropriate human judgment restored" starting point
- Signed independent-verifier declaration — the stance that survives DoD IG or JAG interview
See the lane-shift chart that produces the finding →
$499 Snapshot. 3 business days.
Independent-verifier determination scoped to your program's AI surface + 3 fix-first items + signed declaration.
Buy $499
Snapshot credit applies to Baseline ($2,500) or Enterprise Attestation ($35-55K) upgrade within 30 days.