Use Cases
Command authority, enforced at execution.
Defense organizations already run on delegated authority, classification boundaries, and after-action evidence. Command authority is explicit, bounded, and documented. Access to information is compartmented by construction. Every consequential action is reviewable. That is an accountability model most industries are only now attempting to build, and it is the reason agentic systems in defense fail for a specific reason: not because the model is wrong, but because the software layer cannot express the authority structure the organization actually operates under.
Orcher expresses that model in the agentic layer instead of asking it to bend. Delegation of authority becomes issued scope with an expiry. Compartmentation becomes enforced routing and residency policy checked per call. After-action review becomes a hashed ledger written at execution, not a reconstruction assembled from disparate logs.
The result is agentic capability that mission owners can actually authorize: fast enough to matter operationally, bounded enough to satisfy the authorities that govern it, and evidenced well enough that an inspector general or a coalition partner can verify what happened.
Where liability lands
Delegated command authority and classification control are legal constructs, and software must enforce rather than approximate them. Export control, releasability, and partner-nation restrictions govern where data may be processed, while appropriation and warrant limits bound every contracting action. Orcher issues scoped, expiring authority derived from real delegations, enforces routing by classification per call, and writes tamper-evident records for oversight.
Pressure points
01
Delegated command authority and approval levels are legal constructs. Software must enforce them exactly, not approximate them with policy prompts that a model may or may not respect.
02
Classification and need-to-know boundaries must be enforced at the credential and routing layer. An agent with broad retrieval scope is a spillage event waiting for the right query.
03
Releasability, export control, and partner-nation restrictions vary per data element and per recipient. Provider routing that ignores them creates a disclosure the organization cannot recall.
Named use cases
Real military & defense workflows, each bound to the authority that permits it and reconciled against the systems of record before anything commits.
01
Planning products carry the authorizing role and a verifiable derivation path, so the staff process that produced a course of action can be examined rather than described.
02
Sustainment actions reconcile against readiness, supply, and maintenance systems of record before commit, with obligation authority enforced by role and dollar ceiling.
03
Classification, releasability, and residency are enforced per call, with a full cross-provider trace showing exactly which systems processed which material.
04
Contract actions execute under warranted authority with reconciliation against the contract file and appropriation, producing an audit record the oversight community expects.
05
Actions affecting service members carry the commander's or administrator's authority with verification against personnel systems of record, preserving appeal and redress rights.
06
Response actions are scoped to the responder's authority, verified against network state, and permanently recorded, which is what post-incident review and attribution require.
01 · In depth
The assumption that governance slows operations misreads how defense organizations actually work. Delegated authority exists precisely to enable speed: a commander who knows their limits acts without asking permission.
Orcher applies the same logic to agents. Within issued scope, execution is immediate and unmediated. At the boundary of that scope, the action halts and escalates to a human who holds the authority. Nothing is slowed except the actions that were never authorized in the first place.
02 · In depth
Defense accountability faces two directions: inward toward inspectors general, congressional oversight, and legal review, and outward toward coalition partners who must trust shared products.
A hashed, contemporaneous execution record serves both. Internally it answers who authorized, what was verified, and what was spent. Externally it lets a partner verify the provenance of a product without exposing the underlying sources or the systems that produced it.
Components engaged
These are the components that carry the weight in this industry. Each one is a control, not a recommendation.
Layer 1
Binds the directive to a human and a role, cryptographically and revocably.
Layer 2
Training exclusion and data boundaries verified before routing.
Layer 1
Verifies the proposed action against systems of record before commit.
Layer 1
Hashes the verified action permanently. Evidence, not logs.
Layer 2
Real-time cross-provider trace: which model, which role, what cost, what outcome.
Mechanism
Four stages, in order. Layer 1 components gate execution; Layer 2 components run continuously and never block.
01
The directive is stated and frozen
A commander or authorized official holding the delegation states the outcome in plain language — for example, "Prepare the sustainment plan for this readiness shortfall." It is signed and versioned before any model is called.
02
Authority is minted for this directive only
The Role Identity Fabric resolves the person and their current military & defense role, then mints task-bound, time-bound credentials — median scope around 14% of the underlying account.
03
The proposed action is verified, not reviewed
Before a planning product, logistics commitment, or intelligence work product commits, the Logic Scrubber re-derives the facts it depends on from Authoritative logistics and readiness records, classification markings, rules of engagement, clearance registers. The proposed product mixes material across classification boundaries the directive does not permit — that is a halt, not a warning.
04
The cycle is hashed into the record
The action, the human, the role, the verification and the cost are hashed together. An inspector general, a classification review, or an after-action investigation receives an evidence package, not a reconstruction project.
Operational contract
What this is not
This is not weapons autonomy
Orcher governs administrative, logistics, planning, and analytic agentic work. Delegation of lethal authority is out of scope by design.
Failure behaviour
The proposed product mixes material across classification boundaries the directive does not permit. The directive halts, nothing partial is written, and the halt is recorded with its reason.
Rollout outcomes
Delegation made explicit
Every product resolves to the official whose delegation it exercised.
Classification enforced
Cross-boundary composition halts rather than commits.
After-action evidence
Append-only records survive the investigation they will eventually face.
What the record proves
Orcher writes the proof at execution time. Nothing here depends on reconstructing intent from logs after the fact.
Command-bound
Authority inherited from the chain of command, never assumed
Compartmented
Clearance and need-to-know verified before data moves
Attributable
Every action reconstructable for after-action review
Deployment path
One workflow, one role, one verified execution cycle. Scope widens only after the first cycle holds up under review.
01
Start with mission planning support, logistics, or intelligence workflow steps that already have a named approving authority.
02
Clearance, compartment and command authority come from the identity fabric and cannot be widened downstream.
03
Classification handling, releasability and rules-of-engagement constraints verify before any action executes.
04
After-action review reads a hashed sequence of authorized actions with their verification results attached.
Questions
Direct answers, in the language of the people who carry the consequence.
No. Authority originates with a human in the chain of command, is scoped, and cannot be widened by any agent. The control plane exists to keep that boundary structural.
Compartment and releasability are verified before data moves. A directive that would cross a boundary does not execute.
Routing policy determines what is eligible where. Where verification cannot run, execution stops rather than proceeding unverified.
A per-action record of authority, verification and outcome, produced at execution time rather than reconstructed from logs.
Policy states that a human must be accountable. Orcher makes that accountability a property of execution rather than a statement of intent.
Request a briefing
A working session, not a pitch: your workflow, the role that holds authority for it today, and the seven components that would govern it. Sixty minutes.
Go deeper
The solutions that carry this industry, the research behind the model, and the neighbouring industries with the same accountability problem.
Solution
Authority belongs to a person and a role — never to a service account.
Solution
Verify where data goes before it goes there.
Solution
Turn agent activity into evidence a regulator will accept.
Research
The full thesis: why oversight failed and what replaces it.
Research
How a directive is cryptographically bound to a person and a role.
Research
Announcement and reading guide for the August 2026 volume.
Industry
The same accountability model, applied to government & public sector.
Industry
The same accountability model, applied to logistics & transportation.
Industry
The same accountability model, applied to telecommunications.
Keep reading
Surfaced automatically from the Orcher components this industry relies on.
Layer 1
Binds the directive to a human and a role, cryptographically and revocably.
Layer 2
Training exclusion and data boundaries verified before routing.
Layer 1
Verifies the proposed action against systems of record before commit.
Layer 1
Hashes the verified action permanently. Evidence, not logs.
Layer 2
Real-time cross-provider trace: which model, which role, what cost, what outcome.
Solution
Turn agent activity into evidence a regulator will accept.
Solution
Verify where data goes before it goes there.
Solution
Route by the stakes of the action, not the habits of the developer.
Use case
Physical consequence closes the loop on digital authority.
Use case
Reliability standards do not recognize autonomous intent.
Use case
Professional responsibility is not delegable to a model.
Every deployment starts with one workflow, one role, and one verified execution cycle. Bring the workflow; we will map it to the seven components before you commit to anything.
Case studies
Modelled reference scenarios with the measurement window, the components enforced and the numbers attached. Each one downloads as a PDF.
Military & Defense
Clearance-scoped roles and a fail-closed boundary let controlled programme data reach a model at all — under conditions the programme office could verify.
