ACGM — Assumption Change Governance Module
Architecture Ecosystem: Structured Reality Standards™
Architecture Family: PGA™ — Prediction Governance Architecture
Parent Standard:
Prediction Assumption Standard (PAS)
Operational Layer: Prediction Assumption Governance Layer
Category:
Governance & Enforcement
Subcategory:
Prediction Governance
Type: Parent Standard Module
Governed Space:
Assumption Change Governance
Version: 1.0
Status: Canonical · Open Module
Effective Date: 7 September 2026
Compatibility: OOF® Methodology OS™ · GOA™ · ORA™ · VALIDOS™ · INTEGROS® · ADIT® · AGA™ · AIG® · ASGA™ · CLIA® · SIMULOS®
AI-Readable: Yes
Authority: OOF®
Protection:
MIP® — Methodological Intellectual Property
Canonical Language:
English (UCL™)
Governed Space
Assumption Change Governance
Canonical Definition
Assumption Change Governance Module (ACGM) defines the governance conditions under which changes in Predictive Assumptions are detected, classified, assessed, propagated through relevant dependency structures, associated with their sensitivity and materiality consequences, preserved historically, and translated into governed assumption-state outcomes and prediction-level review signals.
ACGM governs what happens when an assumption that supports a prediction no longer remains in the same governed state.
It transforms:
Governed Assumption Set + Dependency Map + Sensitivity Profile + Materiality Profile
into:
Governed Assumption Change State
and determines whether the change requires:
continued observation; assumption reassessment; qualification; restriction; escalation; prediction-level review trigger.
ACGM does not itself revise, supersede, or withdraw the prediction.
Those actions belong later to:
PMRS — Prediction Monitoring & Revision Standard.
Its foundational distinction is:
Assumption Change ≠ Automatic Prediction Revision
Operational Role
ACGM serves as the fifth and final internal governance space of the Prediction Assumption Standard (PAS) .
It receives:
Identified Predictive Assumptions
Governed Assumption Dependency Map
Governed Assumption Sensitivity Profile
Governed Assumption Materiality Profile
and governs material changes in those assumptions.
The module asks:
What changed in the assumption, how significant is that change, what does it affect, and does the prediction now require broader review?
The governed progression is:
Governed Assumption State → Change Detection → Change Classification → Materiality Reassessment → Dependency Propagation → Assumption State Transition → Historical Preservation → Prediction-Level Review Signal
The primary output is:
Governed Assumption Change Record
and, where applicable:
Prediction Review Trigger
This completes the internal PAS flow and provides the governed assumption structure required by:
PFOS — Prediction Formation Standard
and later by:
PMRS — Prediction Monitoring & Revision Standard.
Internal Module Flow
Predictive Assumption Identification → Assumption Dependency Mapping → Assumption Sensitivity → Assumption Materiality → → Assumption Change Governance
Current Position: Assumption Change Governance
Module Operational Space
ACGM governs:
assumption change detection; assumption-state transition; assumption weakening; assumption strengthening; assumption breach; assumption invalidation; assumption expiry; assumption uncertainty increase; assumption uncertainty reduction; assumption contradiction; assumption restoration; assumption re-entry; partial change; threshold crossing; material change; critical change; dependency propagation; downstream assumption impact; sensitivity reassessment; materiality reassessment; change trigger;
reassessment trigger; monitoring trigger; escalation trigger; prediction-review trigger; change timing; change source; change provenance; temporary change; persistent change; reversible change; irreversible change; historical state preservation; assumption version continuity; retrospective change discovery; and unresolved change state.
Module Function
A prediction may be well governed at formation time.
Its assumptions may be explicit.
Their dependencies may be mapped.
Sensitivity may be understood.
Materiality may be determined.
But the world can change.
A supplier fails.
A regulation changes.
A market regime shifts.
A system configuration changes.
A historical relationship weakens.
A previously stable operating condition disappears.
At that point, assumption governance cannot remain static.
ACGM determines how the assumption state changes and what downstream governance consequences must become visible.
Assumption Change
An Assumption Change occurs when a condition represented by a Predictive Assumption changes materially relative to the governed state under which the prediction was formed or maintained.
Change may affect:
assumption truth state; assumption plausibility; assumption support; assumption scope; assumption boundary; assumption applicability; assumption timing; dependency structure; uncertainty;
sensitivity; materiality.
Change Detection
Change may be detected through:
new Prediction Basis information; Operational Reality updates; monitoring systems; human review; model signals; external events; policy updates; system-state changes; supplier changes; market changes; event occurrence; scheduled reassessment.
Detection does not automatically determine consequence.
Change Detection ≠ Material Change Determination
Assumption-State Transition
A Predictive Assumption may move between governed states.
Potential transitions include:
Holding → Conditionally Holding
Holding → Uncertain
Holding → Under Challenge
Holding → Breached
Conditionally Holding → Holding
Uncertain → Holding
Uncertain → Breached
Breached → Restored
Holding → No Longer Applicable
Unable to Determine → Determined
State transitions must remain historically identifiable.
Assumption Weakening
An assumption may weaken without fully failing.
Example:
A supplier expected to remain operational begins experiencing severe financial stress but continues delivering.
The assumption may move from:
Holding
to:
Conditionally Holding
or:
Under Challenge.
Assumption Weakening ≠ Assumption Breach
Assumption Strengthening
New information may strengthen an assumption.
Example:
A supplier expected to remain operational receives long-term financing and confirms secured production capacity.
Assumption strengthening may reduce uncertainty or change monitoring requirements.
Assumption Breach
An Assumption Breach occurs where the governed condition represented by the assumption no longer holds within its defined boundary.
Example:
Assumption:
The primary production line remains operational during the Prediction Horizon.
Operational Reality:
The production line is shut down indefinitely.
The assumption is breached.
Assumption Breach ≠ Prediction Error
The prediction may still happen to produce the correct outcome through another pathway.
Assumption Invalidation
An assumption may become invalid because its underlying basis, scope, methodology relationship, or governing conditions no longer support its use.
Invalidation may occur even before a clear real-world breach.
Example:
A methodology-dependent continuity assumption becomes unusable after a structural regime change.
Assumption Expiry
Some assumptions are valid only for defined periods.
When temporal applicability ends, the assumption may expire.
Assumption Expiry ≠ Assumption Breach
Expiry reflects the end of governed applicability rather than contradiction.
Assumption Contradiction
New information may directly challenge an assumption.
A contradiction may result in:
uncertainty; challenge state; breach review; materiality reassessment; prediction-review trigger.
Not every contradiction proves the assumption false.
Partial Assumption Change
A compound or bounded assumption may change only partially.
Example:
Supplier capacity remains between 90% and 100%.
Capacity falls to 82%.
The supplier remains operational, but the assumption no longer holds as stated.
ACGM preserves partial change rather than forcing false binary interpretation.
Threshold Crossing
Assumption change may cross a threshold defined under ASNM or AMAM.
Examples:
sensitivity threshold; materiality threshold; criticality threshold; operating-state threshold.
Threshold crossing may trigger stronger governance action.
Threshold Crossing ≠ Automatic Prediction Withdrawal
Material Assumption Change
A Material Assumption Change occurs where assumption-state change may materially alter:
predicted outcome; methodology applicability; probability; uncertainty; feasibility; interpretation; reliance conditions; prediction governability.
Such change requires explicit governance treatment.
Critical Assumption Change
A Critical Assumption Change occurs where a Critical Assumption:
materially weakens; is breached; becomes unable to determine; loses applicability; crosses a critical threshold.
Such changes may require immediate prediction-level review.
Change Classification
ACGM may classify assumption change as:
Immaterial
Minor
Material
Critical
Temporary
Persistent
Reversible
Irreversible
Conditionally Resolved
Unable to Determine
Multiple classifications may apply simultaneously.
Temporary Change
A temporary assumption change may be expected to revert.
Example:
A short-term infrastructure interruption.
Temporary does not automatically mean immaterial.
Temporary Change ≠ Immaterial Change
Persistent Change
A persistent change remains beyond the expected temporary window or establishes a new operating condition.
Persistent change may require assumption redefinition or supersession.
Reversible Change
A reversible change can return to a prior governed state.
Example:
A temporarily unavailable system is restored.
Irreversible Change
An irreversible change permanently alters the assumption context.
Examples:
legal authorization permanently revoked; facility destroyed; supplier permanently exits market; methodology permanently retired.
Irreversibility may materially affect future prediction continuation.
Change Source
Where material, ACGM identifies the source of assumption change.
Possible sources include:
Operational Reality; Prediction Basis update; policy change; human decision; system change; market change; model change; external event; new evidence reference; time expiry.
Change Source ≠ Change Consequence
Change Timing
The timing of assumption change matters.
A change:
before Prediction Formation
may prevent legitimate formation.
A change:
after formation but before reliance
may require re-evaluation.
A change:
after reliance
may affect monitoring and revision.
A change:
after outcome
may matter only for historical interpretation.
Pre-Formation Assumption Change
Where an assumption changes before Prediction Formation, PAS must update the governed assumption structure before PFOS proceeds.
The prediction should not be formed using an assumption state already known to be outdated.
Post-Formation Assumption Change
Where an assumption changes after formation, the original assumption state must remain preserved.
The new state becomes part of continuing governance.
Current Assumption State ≠ Original Assumption State
Dependency Propagation
Assumption change may affect dependent assumptions.
ACGM uses the dependency structure established by ADMM to identify downstream consequences.
Example:
Supplier Insolvency
↓
Supplier Continuity Assumption Breached
↓
Component Availability Assumption Weakens
↓
Production Continuity Assumption Weakens
↓
Output Prediction Requires Review
Propagation Depth
Change may propagate through multiple dependency layers.
A change should not be considered isolated merely because direct predictive impact is initially small.
Propagation Scope
ACGM determines which assumptions or predictive conditions may be affected by a change.
Potential propagation targets include:
other assumptions; Prediction Formation conditions; probability;
uncertainty; feasibility; reliance conditions.
Propagation Containment
Alternative pathways may contain assumption change.
Example:
Primary supplier fails, but qualified backup supplier fully replaces capacity.
The upstream assumption is breached.
Downstream prediction impact may remain limited.
Assumption Breach ≠ Unavoidable Downstream Failure
Sensitivity Reassessment
A changed assumption may alter its sensitivity relationship.
For example:
An assumption near a threshold may move into a region of much higher predictive sensitivity.
ACGM can trigger:
ASNM Reassessment
where necessary.
Materiality Reassessment
Assumption materiality may change as:
dependency structure changes; alternatives disappear; sensitivity changes; Prediction Purpose changes; reliance increases; Prediction Horizon shortens or extends.
ACGM can trigger:
AMAM Reassessment
where necessary.
Change Propagation vs Sensitivity
ADMM establishes:
where change can travel.
ASNM establishes:
how the prediction may respond.
AMAM establishes:
how important that consequence is.
ACGM governs:
what happens when the change actually occurs.
Assumption Reassessment Trigger
A reassessment trigger may include:
new contradictory information; threshold crossing; dependency change; elapsed review interval; operational disruption; methodology change; source deterioration; regulatory change; critical external event.
Monitoring Trigger
A material assumption may require continued observation after a change.
Monitoring may increase where:
uncertainty rises; critical threshold approaches; assumption remains conditionally holding; restoration is expected but unconfirmed.
Escalation Trigger
Assumption change may require escalation where:
critical assumptions are breached; multiple material assumptions change simultaneously; dependency propagation becomes extensive; assumption state cannot be determined; prediction remains in active reliance.
Escalation does not itself determine the final prediction action.
Prediction Review Trigger
A Prediction Review Trigger is a governed signal indicating that assumption change may be sufficiently significant to require broader prediction-level assessment under PMRS.
Potential triggers include:
Critical Assumption Breach; multiple Material Assumption Changes; methodology-applicability loss; feasibility deterioration; major uncertainty expansion; material reliance impact; severe dependency propagation.
Prediction Review Trigger ≠ Prediction Revision Decision
PMRS determines the prediction-level response.
Prediction Formation Relationship
Where material assumption change occurs before Prediction Formation, PFOS must receive the current governed assumption state.
PFOS must not silently consume superseded assumptions.
Probability & Uncertainty Relationship
Assumption change may alter:
probability; uncertainty; confidence; prediction ranges.
PPUS governs the resulting representation.
ACGM provides the relevant assumption-change signal.
Feasibility Relationship
Assumption change may materially affect whether the predicted outcome remains achievable.
PFES governs feasibility determination.
ACGM identifies the assumption change requiring that consideration.
Communication & Reliance Relationship
Where assumption changes materially alter permitted interpretation or reliance, PCRS may require updated communication or reliance conditions.
ACGM does not itself communicate the prediction.
Monitoring & Revision Relationship
PMRS governs:
prediction review; revision; supersession; withdrawal.
ACGM supplies governed assumption-change conditions that may trigger those processes.
Assumption Change Governance ≠ Prediction Revision Governance
Assumption Restoration
A breached or weakened assumption may later be restored.
Example:
A system outage ends and the system returns to its required operating state.
Restoration should create a new governed state.
It must not erase the prior breach.
Restoration ≠ Historical Erasure
Conditional Restoration
An assumption may appear restored but remain subject to conditions.
Example:
Supplier operations resume at only 60% capacity.
The assumption may move from:
Breached
to:
Conditionally Holding
rather than directly to:
Holding.
Assumption Supersession
Where an assumption is replaced by a materially different assumption, the original assumption should remain historically identifiable.
Example:
Original:
Primary supplier remains operational.
New:
Backup supplier provides at least 80% replacement capacity.
These are not the same assumption state.
A new predictive premise has emerged.
Assumption Re-entry
An assumption that ceased to apply may later become applicable again.
Re-entry should be explicitly governed.
Historical Applicability ≠ Current Applicability
Assumption Version Continuity
Material change may require a new assumption version or state record.
The history should preserve:
Initial State → Changed State → Breached State → Restored / Superseded / Retired State
where applicable.
Historical Change Integrity
Later information must not rewrite the assumption history.
A prediction formed under:
Assumption State A
must remain associated with State A even if the assumption later moves to:
State B.
Current Assumption State ≠ Historical Assumption State
Retrospective Change Discovery
A change may be discovered only later.
Example:
An organization learns after two weeks that a supplier had already lost a required authorization five days before the prediction was formed.
The governance record should distinguish:
Actual Change Time
from:
Change Discovery Time
This distinction is essential.
Change Time ≠ Discovery Time
A condition may change before anyone becomes aware of it.
ACGM preserves both times where material.
Unknown Change Time
The exact time of change may sometimes be unknown.
Where so:
Unknown Change Time ≠ No Change
The uncertainty must remain explicit.
Concurrent Assumption Change
Several assumptions may change simultaneously.
This may create:
cumulative effects; interaction effects; dependency cascades; systemic predictive change.
Concurrent changes should not be assessed only as isolated events where their combined effect is material.
Assumption Change Cascade
A change cascade occurs where one assumption change causes or contributes to several downstream state transitions.
Example:
Regulatory Ban
↓
Product Authorization Assumption Breached
↓
Sales Continuity Assumption Breached
↓
Revenue Assumption Breached
↓
Growth Prediction Requires Review
Change Contagion ≠ Independent Failure
Several assumptions changing together may reflect one upstream cause rather than independent failures.
This distinction helps prevent false interpretation of multiple signals as separate events.
Assumption Change Uncertainty
It may be unclear:
whether change occurred; when it occurred; whether it is temporary; whether it is material; whether restoration occurred.
Possible states may include:
Confirmed
Probable
Possible
Disputed
Unable to Determine
Uncertainty itself may justify heightened governance.
Unresolved Assumption Change
Where material change cannot be sufficiently determined, ACGM may produce:
Assumption Change Governance Gap
Possible responses include:
additional monitoring; clarification; restricted reliance; conditional prediction state; escalation; prediction review.
Assumption Change Closure
A change episode may be considered governed when:
change is identified; classification is established; affected dependencies are mapped; sensitivity/materiality consequences are considered; state transition is recorded; required triggers are issued; historical state is preserved.
Closure does not necessarily mean restoration.
Governance Requirements
ACGM requires that:
1. material assumption changes be identifiable; 2. assumption-state transitions remain explicit; 3. weakening remain distinguishable from breach; 4. breach remain distinguishable from prediction error; 5. expiry remain distinguishable from breach; 6. partial changes remain governable; 7. threshold crossing remain explicit; 8. temporary and persistent changes remain distinguishable; 9. reversible and irreversible changes remain distinguishable where material; 10. change source remain identifiable where relevant; 11. change time and discovery time remain distinguishable; 12. dependency propagation be assessed where material; 13. propagation containment remain visible; 14. sensitivity reassessment be triggered where necessary; 15. materiality reassessment be triggered where necessary; 16. material assumption changes remain distinguishable from immaterial changes; 17. Critical Assumption Changes receive appropriate escalation; 18. prediction-review triggers be generated where necessary; 19. prediction-review triggers remain distinguishable from revision decisions; 20. restoration not erase prior breach; 21. assumption supersession preserve prior assumption history; 22. concurrent changes be assessed collectively where necessary; 23. retrospective discoveries preserve actual and discovery timing; 24. unresolved change uncertainty remain explicit; 25. current and historical assumption states remain reconstructable.
Governance Logic
ACGM applies the following governance logic:
Governed Prediction Assumption Set
Governed Assumption Dependency Map
Governed Assumption Sensitivity Profile
Governed Assumption Materiality Profile
↓
Potential Assumption Change Detected
↓
Change Source / Timing / Scope Identified
↓
Assumption State Transition Assessed
↓
Weakening / Breach / Expiry / Invalidation / Restoration Distinguished
↓
Threshold Crossing Assessed
↓
Dependency Propagation Assessed
↓
Sensitivity Reassessment Triggered where necessary
↓
Materiality Reassessment Triggered where necessary
↓
Change Classified
↓
Assumption State Updated
↓
Historical State Preserved
↓
Monitoring / Escalation / Prediction Review Trigger Determined
↓
Governed Assumption Change Record
↓
Current Governed Prediction Assumption Set
Where broader prediction review is required:
Prediction Review Trigger → PMRS — Prediction Monitoring & Revision Standard
Where change remains unresolved:
Assumption Change Governance Gap → Monitoring / Clarification / Qualification / Escalation / Restricted Reliance / Prediction Review
Minimum Implementation Framework
A minimum implementation of ACGM must establish:
A. Assumption Reference
The governed Predictive Assumption affected.
B. Prior Assumption State
The governed state before change.
C. Detected Change
The observed, reported, inferred, or otherwise identified change condition.
D. Change Source
Where materially identifiable:
Prediction Basis / Operational Reality / System Change / Human Decision / Regulation / Market / External Event / Time Expiry / Methodology / Other
E. Change Time
When the underlying condition changed, where known.
F. Discovery Time
When the change became known to the predictive governance process.
G. Change Type
At minimum:
Strengthening / Weakening / Breach / Invalidation / Expiry / Restoration / Supersession / Re-entry / Uncertain Change
H. Change Classification
At minimum:
Immaterial / Minor / Material / Critical / Unable to Determine
I. Dependency Propagation Reference
The relevant downstream dependency relationships affected.
J. Sensitivity Reassessment State
Whether ASNM reassessment is required.
K. Materiality Reassessment State
Whether AMAM reassessment is required.
L. New Assumption State
The governed state after change assessment.
M. Monitoring / Escalation State
Any required observation, escalation, or restriction.
N. Prediction Review Trigger
At minimum:
Not Required / Required / Conditional / Unable to Determine
O. Historical Preservation
The prior assumption state, transition, timing, and change record must remain reconstructable.
Governance Outputs
ACGM may produce:
Governed Assumption Change Record; Assumption State Transition Record; assumption-change classification; assumption-weakening record; Assumption Breach Record; Assumption Expiry Record; Assumption Invalidation Record; Assumption Restoration Record; assumption-supersession record; assumption-re-entry record; threshold-crossing record; dependency-propagation record; sensitivity-reassessment trigger; materiality-reassessment trigger; assumption-monitoring trigger; escalation trigger; Prediction Review Trigger; current Governed Prediction Assumption Set; historical assumption-state record; change-time record; discovery-time record; retrospective-change record; unresolved change requirements.
The primary governed output is:
Governed Assumption Change Record
with the updated:
Governed Prediction Assumption Set
and, where materially required:
Prediction Review Trigger
Critical Distinctions
Assumption Change ≠ Automatic Prediction Revision
Change Detection ≠ Material Change Determination
Assumption Weakening ≠ Assumption Breach
Assumption Breach ≠ Prediction Error
Assumption Expiry ≠ Assumption Breach
Threshold Crossing ≠ Automatic Prediction Withdrawal
Temporary Change ≠ Immaterial Change
Change Source ≠ Change Consequence
Current Assumption State ≠ Original Assumption State
Assumption Breach ≠ Unavoidable Downstream Failure
Prediction Review Trigger ≠ Prediction Revision Decision
Assumption Change Governance ≠ Prediction Revision Governance
Restoration ≠ Historical Erasure
Historical Applicability ≠ Current Applicability
Change Time ≠ Discovery Time
Unknown Change Time ≠ No Change
Current Assumption State ≠ Historical Assumption State
Change Contagion ≠ Independent Failure
Enterprise Application
ACGM enables organizations to maintain prediction governance after the moment of Prediction Formation.
Without Assumption Change Governance, a prediction may continue to be:
displayed; communicated; relied upon; automated; operationalized
even after one of its fundamental predictive conditions has materially changed.
This is particularly important in:
AI forecasting + supply-chain planning + finance + insurance + infrastructure + industrial systems + healthcare + strategic planning + market prediction + autonomous systems.
ACGM creates the bridge between:
assumption governance
and:
prediction monitoring and revision.
It ensures that prediction review can be triggered by changes in the conditions that made the prediction meaningful in the first place.
Use Case 1 — The Prediction Is Still Active After Its Critical Supplier Assumption Has Failed
A manufacturer predicts that output will remain above:
50,000 units per month
for the next six months.
One Critical Assumption states:
The sole supplier of a specialized component remains operational and capable of delivering minimum required volume.
Two months later, the supplier enters insolvency and stops production.
The prediction dashboard continues displaying the original forecast because no mechanism links supplier failure to the assumption structure.
The critical governance question is:
Can a prediction remain legitimately active when a Critical Assumption has already been breached but the prediction has no governed mechanism for detecting and propagating that change?
ACGM records:
Assumption State: Holding → Breached
maps the downstream dependency effects,
triggers sensitivity and materiality reassessment where necessary,
and produces:
Prediction Review Trigger: Required
for PMRS.
It does not itself rewrite the prediction.
Assumption Change ≠ Automatic Prediction Revision
Use Case 2 — The Market Forecast Looks Unchanged, but Its Historical Assumption No Longer Holds
An AI system predicts continued customer demand based on several years of stable purchasing behavior.
A core assumption states:
The historical relationship between price, customer behavior, and demand remains sufficiently applicable throughout the Prediction Horizon.
A new regulation fundamentally changes pricing incentives.
Demand has not yet visibly collapsed.
The central forecast remains nearly unchanged.
However, the continuity assumption underlying the predictive methodology is now materially under challenge.
The critical governance question is:
Should a prediction continue to be relied upon until its output visibly fails, or should material change in the assumptions supporting it trigger governance before the error appears?
ACGM moves the assumption from:
Holding
to:
Under Challenge
preserves the regulatory change and timing,
initiates relevant sensitivity and materiality reassessment,
and may generate a Prediction Review Trigger.
Prediction governance should react to material changes in predictive conditions before those changes become historical prediction errors.
Canonical Boundary
ACGM governs the detection, classification, state transition, breach, expiry, invalidation, restoration, dependency propagation, reassessment triggering, escalation, prediction-review signaling, version continuity, timing, and historical preservation of material changes in Predictive Assumptions. It does not govern initial assumption identification, dependency mapping itself, sensitivity assessment itself, materiality determination itself, Prediction Formation, probability representation, feasibility determination, communication and reliance, Operational Reality, or the prediction-level decision to revise, supersede, or withdraw a prediction.
ACGM establishes:
what has changed in the assumptions supporting the prediction and what broader governance must now be triggered.
It does not determine:
the final new prediction state.
Canonical Closing Statement
A governed assumption cannot become a permanent frozen condition inside a changing world.
ACGM ensures that when the conditions beneath a prediction change, the predictive governance structure changes with them — without erasing what was assumed before.
A prediction should not have to fail before governance notices that the assumptions supporting it already have.