Archetypal field notes

Research & insights

Design study / Guarded

Prompt policy review: design a usable human review

An aerial view of a highway interchange beside a waterway in Belmont, California.

All articles

Guarded · Field note 001

Civilian and Air Force software leaders discuss projects around a table with open laptops.

Guarded review handover: define the mission boundary

Guarded · Field note 002

Emergency operations personnel coordinate around tables and laptops beneath wall-mounted status screens.

Guarded review handover: preserve source provenance

Guarded · Field note 003

Civilian developers and a uniformed soldier work together at computers and equipment in a technical workshop.

Guarded review handover: translate policy into a decision

Guarded · Field note 004

Civilian emergency responders and Air National Guard medics review a clipboard together outdoors.

Guarded review handover: design a usable human review

Guarded · Field note 005

A civilian software engineer discusses a project with an Air Force product manager at a shared computer desk.

Guarded review handover: keep permissions bounded

Guarded · Field note 006

Three Defense Logistics Agency technology colleagues review a hand-drawn systems diagram and project calendar.

Guarded review handover: make uncertainty actionable

Guarded · Field note 007

Two Navy network engineering professionals review a rack drawing beside server equipment.

Guarded review handover: test the difficult boundaries

Guarded · Field note 008

A civilian software engineer and Air Force colleagues in casual clothing discuss a software product beside shared workstations.

Guarded review handover: build a useful decision record

Guarded · Field note 009

Civilian and uniformed cyber specialists take part in a collaborative tabletop exercise in a wood-paneled planning room.

Guarded review handover: control change after deployment

Guarded · Field note 010

Civilian emergency operations staff and Space Launch Delta 45 personnel discuss procedures around a conference table.

Guarded review handover: learn from tested outcomes

Policy Studio · Field note 011

An Air Force emergency management specialist and a civilian communications supervisor test radios inside a mobile operations center.

Policy exception handling: define the mission boundary

Policy Studio · Field note 012

A civilian technology lead demonstrates an augmented-reality headset under a field canopy.

Policy exception handling: preserve source provenance

JavaScript enables search and pagination. All field notes are listed below.

Guarded review handover: define the mission boundary

Guarded review handover: preserve source provenance

Guarded review handover: translate policy into a decision

Guarded review handover: design a usable human review

Guarded review handover: keep permissions bounded

Guarded review handover: make uncertainty actionable

Guarded review handover: test the difficult boundaries

Guarded review handover: build a useful decision record

Guarded review handover: control change after deployment

Guarded review handover: learn from tested outcomes

Policy exception handling: define the mission boundary

Policy exception handling: preserve source provenance

Policy exception handling: translate policy into a decision

Policy exception handling: design a usable human review

Policy exception handling: keep permissions bounded

Policy exception handling: make uncertainty actionable

Policy exception handling: test the difficult boundaries

Policy exception handling: build a useful decision record

Policy exception handling: control change after deployment

Policy exception handling: learn from tested outcomes

Managed endpoint governance: define the mission boundary

Managed endpoint governance: preserve source provenance

Managed endpoint governance: translate policy into a decision

Managed endpoint governance: design a usable human review

Managed endpoint governance: keep permissions bounded

Managed endpoint governance: make uncertainty actionable

Managed endpoint governance: test the difficult boundaries

Managed endpoint governance: build a useful decision record

Managed endpoint governance: control change after deployment

Managed endpoint governance: learn from tested outcomes

Decision evidence release: define the mission boundary

Decision evidence release: preserve source provenance

Decision evidence release: translate policy into a decision

Decision evidence release: design a usable human review

Decision evidence release: keep permissions bounded

Decision evidence release: make uncertainty actionable

Decision evidence release: test the difficult boundaries

Decision evidence release: build a useful decision record

Decision evidence release: control change after deployment

Decision evidence release: learn from tested outcomes

Guarded Chat review: define the mission boundary

Guarded Chat review: preserve source provenance

Guarded Chat review: translate policy into a decision

Guarded Chat review: design a usable human review

Guarded Chat review: keep permissions bounded

Guarded Chat review: make uncertainty actionable

Guarded Chat review: test the difficult boundaries

Guarded Chat review: build a useful decision record

Guarded Chat review: control change after deployment

Guarded Chat review: learn from tested outcomes

Critical infrastructure review: define the mission boundary

Critical infrastructure review: preserve source provenance

Critical infrastructure review: translate policy into a decision

Critical infrastructure review: design a usable human review

Critical infrastructure review: keep permissions bounded

Critical infrastructure review: make uncertainty actionable

Critical infrastructure review: test the difficult boundaries

Critical infrastructure review: build a useful decision record

Critical infrastructure review: control change after deployment

Critical infrastructure review: learn from tested outcomes

Policy-pack rollout: define the mission boundary

Policy-pack rollout: preserve source provenance

Policy-pack rollout: translate policy into a decision

Policy-pack rollout: design a usable human review

Policy-pack rollout: keep permissions bounded

Policy-pack rollout: make uncertainty actionable

Policy-pack rollout: test the difficult boundaries

Policy-pack rollout: build a useful decision record

Policy-pack rollout: control change after deployment

Policy-pack rollout: learn from tested outcomes

Disconnected Guarded operation: define the mission boundary

Disconnected Guarded operation: preserve source provenance

Disconnected Guarded operation: translate policy into a decision

Disconnected Guarded operation: design a usable human review

Disconnected Guarded operation: keep permissions bounded

Disconnected Guarded operation: make uncertainty actionable

Disconnected Guarded operation: test the difficult boundaries

Disconnected Guarded operation: build a useful decision record

Disconnected Guarded operation: control change after deployment

Disconnected Guarded operation: learn from tested outcomes

Model change evaluation: define the mission boundary

Model change evaluation: preserve source provenance

Model change evaluation: translate policy into a decision

Model change evaluation: design a usable human review

Model change evaluation: keep permissions bounded

Model change evaluation: make uncertainty actionable

Model change evaluation: test the difficult boundaries

Model change evaluation: build a useful decision record

Model change evaluation: control change after deployment

Model change evaluation: learn from tested outcomes

Evidence-grounded briefing: define the mission boundary

Evidence-grounded briefing: preserve source provenance

Evidence-grounded briefing: translate policy into a decision

Evidence-grounded briefing: design a usable human review

Evidence-grounded briefing: keep permissions bounded

Evidence-grounded briefing: make uncertainty actionable

Evidence-grounded briefing: test the difficult boundaries

Evidence-grounded briefing: build a useful decision record

Evidence-grounded briefing: control change after deployment

Evidence-grounded briefing: learn from tested outcomes

Interpreting assurance scores: define the mission boundary

Interpreting assurance scores: preserve source provenance

Interpreting assurance scores: translate policy into a decision

Interpreting assurance scores: design a usable human review

Interpreting assurance scores: keep permissions bounded

Interpreting assurance scores: make uncertainty actionable

Interpreting assurance scores: test the difficult boundaries

Interpreting assurance scores: build a useful decision record

Interpreting assurance scores: control change after deployment

Interpreting assurance scores: learn from tested outcomes

Human–AI oversight research: define the mission boundary

Human–AI oversight research: preserve source provenance

Human–AI oversight research: translate policy into a decision

Human–AI oversight research: design a usable human review

Human–AI oversight research: keep permissions bounded

Human–AI oversight research: make uncertainty actionable

Human–AI oversight research: test the difficult boundaries

Human–AI oversight research: build a useful decision record

Human–AI oversight research: control change after deployment

Human–AI oversight research: learn from tested outcomes

Decision lineage analysis: define the mission boundary

Decision lineage analysis: preserve source provenance

Decision lineage analysis: translate policy into a decision

Decision lineage analysis: design a usable human review

Decision lineage analysis: keep permissions bounded

Decision lineage analysis: make uncertainty actionable

Decision lineage analysis: test the difficult boundaries

Decision lineage analysis: build a useful decision record

Decision lineage analysis: control change after deployment

Decision lineage analysis: learn from tested outcomes

Case review continuity: define the mission boundary

Case review continuity: preserve source provenance

Case review continuity: translate policy into a decision

Case review continuity: design a usable human review

Case review continuity: keep permissions bounded

Case review continuity: make uncertainty actionable

Case review continuity: test the difficult boundaries

Case review continuity: build a useful decision record

Case review continuity: control change after deployment

Case review continuity: learn from tested outcomes

Guarded deployment changes: define the mission boundary

Guarded deployment changes: preserve source provenance

Guarded deployment changes: translate policy into a decision

Guarded deployment changes: design a usable human review

Guarded deployment changes: keep permissions bounded

Guarded deployment changes: make uncertainty actionable

Guarded deployment changes: test the difficult boundaries

Guarded deployment changes: build a useful decision record

Guarded deployment changes: control change after deployment

Guarded deployment changes: learn from tested outcomes

Cross-team evidence exchange: define the mission boundary

Cross-team evidence exchange: preserve source provenance

Cross-team evidence exchange: translate policy into a decision

Cross-team evidence exchange: design a usable human review

Cross-team evidence exchange: keep permissions bounded

Cross-team evidence exchange: make uncertainty actionable

Cross-team evidence exchange: test the difficult boundaries

Cross-team evidence exchange: build a useful decision record

Cross-team evidence exchange: control change after deployment

Cross-team evidence exchange: learn from tested outcomes

Unmanaged AI use: define the mission boundary

Unmanaged AI use: preserve source provenance

Unmanaged AI use: translate policy into a decision

Unmanaged AI use: design a usable human review

Unmanaged AI use: keep permissions bounded

Unmanaged AI use: make uncertainty actionable

Unmanaged AI use: test the difficult boundaries

Unmanaged AI use: build a useful decision record

Unmanaged AI use: control change after deployment

Unmanaged AI use: learn from tested outcomes

Prompt policy review: define the mission boundary

Prompt policy review: preserve source provenance

Prompt policy review: translate policy into a decision

Prompt policy review: design a usable human review

Prompt policy review: keep permissions bounded

Prompt policy review: make uncertainty actionable

Prompt policy review: test the difficult boundaries

Prompt policy review: build a useful decision record

Prompt policy review: control change after deployment

Prompt policy review: learn from tested outcomes

Agent permission boundaries: define the mission boundary

Agent permission boundaries: preserve source provenance

Agent permission boundaries: translate policy into a decision

Agent permission boundaries: design a usable human review

Agent permission boundaries: keep permissions bounded

Agent permission boundaries: make uncertainty actionable

Agent permission boundaries: test the difficult boundaries

Agent permission boundaries: build a useful decision record

Agent permission boundaries: control change after deployment

Agent permission boundaries: learn from tested outcomes

Precedent applicability: define the mission boundary

Precedent applicability: preserve source provenance

Precedent applicability: translate policy into a decision

Precedent applicability: design a usable human review

Precedent applicability: keep permissions bounded

Precedent applicability: make uncertainty actionable

Precedent applicability: test the difficult boundaries

Precedent applicability: build a useful decision record

Precedent applicability: control change after deployment

Precedent applicability: learn from tested outcomes

Frontier-model evaluation: define the mission boundary

Frontier-model evaluation: preserve source provenance

Frontier-model evaluation: translate policy into a decision

Frontier-model evaluation: design a usable human review

Frontier-model evaluation: keep permissions bounded

Frontier-model evaluation: make uncertainty actionable

Frontier-model evaluation: test the difficult boundaries

Frontier-model evaluation: build a useful decision record

Frontier-model evaluation: control change after deployment

Frontier-model evaluation: learn from tested outcomes

Policy compilation research: define the mission boundary

Policy compilation research: preserve source provenance

Policy compilation research: translate policy into a decision

Policy compilation research: design a usable human review

Policy compilation research: keep permissions bounded

Policy compilation research: make uncertainty actionable

Policy compilation research: test the difficult boundaries

Policy compilation research: build a useful decision record

Policy compilation research: control change after deployment

Policy compilation research: learn from tested outcomes

Research-to-deployment transition: define the mission boundary

Research-to-deployment transition: preserve source provenance

Research-to-deployment transition: translate policy into a decision

Research-to-deployment transition: design a usable human review

Research-to-deployment transition: keep permissions bounded

Research-to-deployment transition: make uncertainty actionable

Research-to-deployment transition: test the difficult boundaries

Research-to-deployment transition: build a useful decision record

Research-to-deployment transition: control change after deployment

Research-to-deployment transition: learn from tested outcomes

Cross-model continuity: define the mission boundary

Cross-model continuity: preserve source provenance

Cross-model continuity: translate policy into a decision

Cross-model continuity: design a usable human review

Cross-model continuity: keep permissions bounded

Cross-model continuity: make uncertainty actionable

Cross-model continuity: test the difficult boundaries

Cross-model continuity: build a useful decision record

Cross-model continuity: control change after deployment

Cross-model continuity: learn from tested outcomes

Guarded

Explore Guarded

Capabilities

Policy Studio

Explore Policy Studio

Capabilities

MissionSAFE

Explore MissionSAFE

Capabilities

Evidence

Explore Evidence

Capabilities

Archetypal AI

Company

Work with us

Research & insights

Prompt policy review: design a usable human review

All articles

Guarded review handover: define the mission boundary

Guarded review handover: preserve source provenance

Guarded review handover: translate policy into a decision

Guarded review handover: design a usable human review

Guarded review handover: keep permissions bounded

Guarded review handover: make uncertainty actionable

Guarded review handover: test the difficult boundaries

Guarded review handover: build a useful decision record

Guarded review handover: control change after deployment

Guarded review handover: learn from tested outcomes

Policy exception handling: define the mission boundary

Policy exception handling: preserve source provenance