HSCM — Harm State Classification Module
OOF® Origin Open Foundation®
Independent Methodological Authority
Canonical Definition
The HSCM — Harm State Classification Module governs how a Governed HarmState Identification is evaluated and assigned to the canonical Harm State
that most accurately represents the current governed condition of Harm.
HSCM establishes consistent classification across changing, intermittent,
distributed, composite, collective, and multi-dimensional Harm while
preserving material uncertainty, qualifications, subject-level
differences, and state boundaries.
It establishes the transition:
Governed Harm State Identification → Classification Basis → State
Eligibility → State Differentiation → State Assignment → Governed Harm
State Classification
Its central question is:
Which canonical Harm State most accurately represents the currently
identified condition of this governed Harm?
- Operational Role
HSIM identifies the current condition.
HSCM determines what that condition methodologically means within the HSS
State Model.
Its operational role is:
to convert an identified Harm condition into a consistent canonical Harm
State without confusing severity, visibility, mechanism status,
intervention activity, prediction, or administrative outcome with the
actual lifecycle state of Harm.
- Position Within HSS
The HSS governance flow is:
State Identification → State Classification → State Persistence → State
Transition → State Monitoring
The Core Module sequence is:
HSIM → HSCM → HSTM → HSPM → HSVM
HSCM receives:
Governed Harm State Identification
from HSIM.It produces:
Governed Harm State Classification
for subsequent HSS governance.- Canonical State Model
HSCM uses eight canonical Harm States:
Emerging
Active
Escalating
Stabilized
Reduced
Recovering
Residual
Resolved
These states represent the governed condition of Harm.They are not:
severity levels
risk ratings
confidence scores
legal findings
claim statuses
or:
system-operational statuses
- Foundational Classification Principle
The foundational rule is:
Harm must be classified according to its current governed condition, not
according to the status of the mechanism, intervention, organization,
claim, or system surrounding it.
Therefore:
Context Status ≠ Harm State
- Classification Basis
State Classification must be derived from the Governed Harm State
Identification produced by HSIM.
The classification basis should preserve, where materially relevant:
current manifestation
persistence
direction of change
affected scope
temporal behavior
subject condition
effect condition
recovery condition
residual condition
uncertainty
and:
material limitations
HSCM must not manufacture information missing from HSIM.
- Emerging
Emerging applies where the governed Harm already exists but its current
harmful condition remains materially in formation, development, or
manifestation.
It may apply where:
effects are developing
extent remains unfolding
manifestation is increasing in definition
or:
the Harm condition has not yet reached a sufficiently established
operational pattern
The distinction is:
Emerging Harm State ≠ Harm Emergence Determination
HEMS governs how Harm begins to exist.
HSCM classifies the current state of an already governed Harm.
- Active
Active applies where Harm is materially present and operative within its
governed boundary without sufficient basis for a more specific lifecycle
classification.
Active Harm may be:
continuous
intermittent
stable
variable
visible
or:
partially latent
Canonical distinction:
Active ≠ Escalating
Harm does not need to worsen in order to remain Active.
- Escalating
Escalating applies where the existing Harm is materially worsening.
Worsening may involve:
greater intensity
greater functional consequence
increasing persistence
increasing affected scope within the existing boundary
or another material deterioration.
However:
Escalating ≠ Propagating
Harm may worsen without spreading to new subjects, effects, systems, or
boundaries.
Propagation dynamics belong to HCPS.
- Stabilized
Stabilized applies where the Harm has ceased materially worsening or
improving over the relevant observation period and maintains a
sufficiently stable harmful condition.
Stabilized does not mean harmless.
Therefore:
Stabilized ≠ Resolved
and:
Stabilized ≠ Recovered
A severe Harm may remain Stabilized for a long period.
- Reduced
Reduced applies where the existing Harm has materially decreased relative
to a prior governed condition.
Reduction may concern:
intensity
effect
scope
persistence
or another governed Harm dimension.
Reduction does not require complete Recovery.
Therefore:
Reduced ≠ Recovered
and:
Reduced ≠ Resolved
- Recovering
Recovering applies where a governed recovery process or materially
supportable restoration trajectory is underway and the Harm condition is
moving toward restoration while Harm remains relevant.
Recovering must not be assigned merely because:
treatment started
repair started
compensation was paid
mitigation occurred
or:
a recovery plan exists
Canonical distinction:
Recovery Activity ≠ Recovering Harm State
The actual Harm condition must support the classification.
- Residual
Residual applies where material Harm remains after recovery, restoration,
mitigation, cessation of the producing mechanism, or another substantial
reduction process has occurred.
Residual Harm may be:
permanent
long-term
latent
intermittent
irreversible
or:
capable of later change
Canonical principle:
What remains after Recovery is still Harm where the governed Harm
condition persists.
- Resolved
Resolved applies only where the applicable Harm Resolution conditions have
been satisfied.
HSCM may represent a Resolved state resulting from appropriate downstream
HAGA™ governance, but it must not manufacture Resolution merely from:
lack of complaint
lack of observation
system repair
claim closure
compensation
mitigation
temporary improvement
or:
mechanism termination
And:
Resolved ≠ Finality
Finality belongs to HRFS.
- Classification Is Not Necessarily Linear
The canonical State Model is not a mandatory sequence.
Harm may transition:
Emerging → Active
Active → Escalating
Escalating → Stabilized
Stabilized → Reduced
but it may also transition:
Reduced → Active
Recovering → Escalating
Residual → Active
or through other materially supported paths.
Canonical principle:
Harm State classification represents reality; reality is not required to
follow a predetermined state sequence.
- Classification Precision
HSCM should assign the most precise state supported by the governed basis.
It must not select a more specific classification where the basis only
supports a broader one.
For example, where Harm is clearly present but whether it is worsening
remains uncertain:
Active / Escalation Uncertain
may be more legitimate than unsupported:Escalating
Canonical principle:Classification precision must not exceed governance precision.
- Primary State and Qualifications
Complex Harm may require:
Primary Harm State + Material Qualifications
For example:
Active — Intermittent
Escalating — Subject Scope Partially Uncertain
Reduced — Residual Effect Assessment Pending
This allows HSCM to preserve complexity without creating unnecessary newcanonical states.
- Multi-Dimensional Classification
Different Harm dimensions may legitimately occupy different conditions.
For example:
Physical Dimension — Reduced
Financial Dimension — Active
Operational Dimension — Stabilized
HSCM should preserve such differences where collapsing them into one labelwould materially distort the Harm.
A primary classification may still be used where methodologically
justified.
- Subject-Level Classification
Where multiple subjects experience materially different Harm conditions,
HSCM may preserve subject-level classifications.
For example:
Subject A — Recovering
Subject B — Active
Subject C — Residual
Canonical principle:One Subject's State ≠ Whole-Harm State by Default
- Collective and Composite Classification
Collective or composite Harm must be classified at the level at which the
Harm actually exists.
Therefore:
Majority Improvement ≠ Collective Harm Resolution
and:
Component Resolution ≠ Composite Harm Resolution
HSCM must avoid averaging away materially continuing Harm.
- Distributed Harm Classification
Where Harm spans multiple systems, agents, locations, organizations, or
domains, technical fragmentation must not dictate classification.
Canonical principle:
Distributed Harm requires governed classification across the material Harm
object, not separate assumptions based on system boundaries.
- Intermittent Harm
Harm that temporarily becomes unobservable or inactive in manifestation
may remain governed.
HSCM must not repeatedly classify:
Active → Resolved → Active → Resolved
merely because an intermittent manifestation disappears and returns.
Classification must reflect the underlying Harm condition and relevant
temporal pattern.
Therefore:
Temporary Non-Manifestation ≠ Resolution
- Severity Boundary
Harm severity and Harm State are separate dimensions.
A Harm may be:
Severe + Stabilized
Moderate + Escalating
Low + Active
Severe + Residual
Therefore:
Severity ≠ State
HSCM must not use severity labels as substitutes for lifecycle
classification.
- Direction of Change
Direction of change is important for distinguishing states.
Conceptually:
Developing → Emerging
Present without material directional basis → Active
Worsening → Escalating
Materially stable → Stabilized
Improving / diminished → Reduced
Restoration underway → Recovering
Remaining after recovery → Residual
Resolution conditions satisfied → Resolved
This mapping is methodological guidance, not an automatic algorithm
detached from the governed Harm context.
- Mechanism and System Boundary
The state of the Harm-producing mechanism cannot determine Harm State by
itself.
Therefore:
System Active ≠ Harm Active
System Stopped ≠ Harm Resolved
System Fixed ≠ Harm Recovered
Model Removed ≠ Harm Removed
HSCM classifies the Harm consequence, not the mechanism.
- Intervention Boundary
Containment and Mitigation actions may affect Harm State but do not define
it.
Therefore:
Containment Applied ≠ Stabilized
Mitigation Applied ≠ Reduced
Recovery Program Started ≠ Recovering
The actual governed Harm condition must support the resulting
classification.
- Accountability, Liability and Claim Boundary
Harm State classification must remain independent from:
accountability
responsibility
liability
claim acceptance
claim rejection
compensation
or:
settlement
Therefore:
Claim Closed ≠ Harm Resolved
Compensation Paid ≠ Harm Recovered
No Liability ≠ No Active Harm
- Prediction, Simulation and Risk Boundary
A system may predict that Harm will escalate.
SIMULOS® may simulate escalation.
Risk governance may identify high probability of worsening.
But:
Predicted Escalation ≠ Escalating State
Simulated Recovery ≠ Recovering State
Future Risk ≠ Current Harm State
HSCM classifies actual governed Harm conditions.
- Uncertainty and Conflict
Where HSIM provides material uncertainty or conflicting current-condition
information, HSCM must preserve it.
Possible outcomes include:
classification with qualification
classification deferred
or:
classification reassessment required
HSCM must not resolve uncertainty by selecting whichever state is
operationally most convenient.
Canonical principle:
Uncertainty must qualify classification, not disappear inside it.
- Classification Confidence
A classification may carry an associated confidence where useful.
Confidence may be:
high
moderate
low
or another governed representation.
But:
Classification Confidence ≠ Harm State
and:
High Confidence ≠ High Severity
- Classification Readiness
HSCM should classify only where HSIM provides a sufficiently governed
current-condition basis.
If the current condition is materially unidentified, temporally
misaligned, or fundamentally conflicted:
→ classification should not be fabricated.
Canonical principle:
No Classification is preferable to false classification where the current
Harm condition is not sufficiently governable.
- Reclassification
A Harm may be reclassified when new governed information establishes that
its current condition has changed or a previous classification was
incorrect.
Reclassification must distinguish:
State Transition
from:Classification Correction
A transition means the Harm changed.A correction means the previous representation of the Harm was wrong.
These are not the same governance event.
- Classification History
HSCM must preserve sufficient history to distinguish:
what state was assigned
when it applied
what basis supported it
whether it later transitioned
and:
whether it was later corrected
Canonical principle:
Current Classification ≠ Complete Classification History
- Handoff to Subsequent HSS Governance
HSCM produces:
Governed Harm State Classification
The output should preserve at minimum:Harm identity
canonical Harm Stateapplicable time
scope
material qualifications
dimensional or subject-level distinctions where necessary
uncertainty
classification confidence where used
classification basis reference
and:
classification status
This classification becomes the governed basis for subsequent State
Persistence, Transition, and Monitoring governance.
- Minimum Implementation Framework
A minimum HSCM implementation requires eight capabilities.
- Canonical State Model
Support Emerging, Active, Escalating, Stabilized, Reduced, Recovering,
Residual, and Resolved as semantically distinct states.
- Classification Basis Control
Classify from the Governed Harm State Identification rather than
mechanism, administrative, claim, or intervention status.
- State Differentiation
Distinguish materially adjacent states, particularly Active vs Escalating,
Stabilized vs Reduced, Reduced vs Recovering, and Residual vs Resolved.
- Complex Harm Representation
Support qualified, multi-dimensional, subject-level, collective,
composite, distributed, and intermittent Harm classification.
- Uncertainty Governance
Preserve uncertainty, conflict, classification limitations, and confidence
without forcing false precision.
- Boundary Preservation
Keep severity, risk, prediction, simulation, operational state,
accountability, liability, claims, and intervention status separate from
Harm State.
- Reclassification & History
Distinguish genuine State Transition from Classification Correction and
preserve historical classification continuity.
- Governed Output
Produce a traceable Governed Harm State Classification for subsequent HSS
governance. 36. Use Case 1 — System Repaired, Harm Remains
An AI-controlled operational system causes a physical Harm.
The technical defect is repaired and the system returns to normal
operation.
The affected person's Harm remains materially present but is no longer
worsening.
HSCM does not classify the Harm as:
Resolved
because the system was repaired.Based on the actual Harm condition, the appropriate classification may be:
Stabilized
The governing principle is:Restoration of the producing system does not establish restoration of the
harmed subject.
- Use Case 2 — Composite Harm Improves Unevenly
A distributed AI operation produces a governed Harm spanning:
digital
financial
and:
organizational
effects.
The digital effect is removed.
The financial Harm remains Active.
The organizational consequence is Recovering.
HSCM preserves these material differences rather than declaring the entire
Harm:
Resolved
because one component disappeared.The governing principle is:
A composite Harm must not be classified by the condition of its most
improved component.
- Governance Requirements
HSCM requires that:
- classification derive from a Governed Harm State Identification
- all eight canonical Harm States remain semantically distinct
- Emerging State remain separate from Harm Emergence governance
- Active remain distinguishable from Escalating
- Escalation remain distinguishable from Propagation
- Stabilized remain distinguishable from Reduced and Resolved
- Reduced remain distinguishable from Recovering
- Residual Harm remain distinguishable from Resolution
- Resolved remain distinguishable from Finality
- classification precision not exceed the governed basis
- complex Harm permit qualified and dimensional classification
- subject, collective, composite, and distributed differences
remain
representable; - intermittent non-manifestation not automatically become
Resolution; - severity remain separate from lifecycle State; - mechanism
and intervention status not determine State; - accountability, liability,
claims, and compensation remain separate; - predictions, simulations, and
future risk not substitute for current State; - uncertainty remain
visible; - false classification not be forced where the basis is
insufficient; - State Transition remain distinguishable from
Classification Correction; - classification history remain traceable.
- Governance Logic
HSCM applies the following logic:
Has HSIM produced a sufficiently governed Harm State Identification?
If no:
→ do not fabricate a canonical classification.
If yes:
→ evaluate the identified current Harm condition.
Is Harm still materially developing in its manifestation?
→ consider Emerging.
Is Harm materially present without sufficient basis for another more
specific state?
→ consider Active.
Is existing Harm materially worsening?
→ consider Escalating.
Has Harm become materially stable without sufficient reduction?
→ consider Stabilized.
Has Harm materially decreased?
→ consider Reduced.
Is actual recovery materially underway?
→ consider Recovering.
Does material Harm remain following recovery or substantial restoration?
→ consider Residual.
Have applicable Harm Resolution conditions been legitimately satisfied?
→ represent Resolved, without treating it as Finality.
Then test:
Is the classification being inferred from mechanism status, intervention,
claim status, prediction, simulation, severity, or administrative
convenience rather than Harm condition?
If yes:
→ reject or reassess the classification basis.
Are different subjects, effects, or dimensions materially different?
If yes:
→ preserve the relevant qualifications.
Is uncertainty material?
If yes:
→ preserve it.
Does new information indicate genuine Harm change or previous
classification error?
→ distinguish State Transition from Classification Correction.
Then produce:
Governed Harm State Classification
At every stage:Severity ≠ State
Mechanism Status ≠ Harm State
Intervention ≠ State
Prediction ≠ Current State
Temporary Non-Manifestation ≠ Resolution
Resolved ≠ Finality
- Governance Outputs
HSCM may produce:
Governed Harm State Classification
Canonical Harm State
Emerging Classification
Active Classification
Escalating Classification
Stabilized Classification
Reduced Classification
Recovering Classification
Residual Classification
Resolved Classification
State Qualification
Dimensional State Classification
Subject-Level State Classification
Collective State Classification
Composite State Classification
Distributed State Classification
Intermittent Harm Qualification
Classification Uncertainty
Classification Confidence
Classification Limitation
Classification Correction
Classification History
The principal output is:Governed Harm State Classification
- Critical Distinctions
State Identification ≠ State Classification
Emerging State ≠ Harm Emergence Determination
Active ≠ Escalating
Escalation ≠ Propagation
Stabilized ≠ Resolved
Reduced ≠ Recovered
Reduced ≠ Resolved
Recovery Activity ≠ Recovering State
Residual ≠ Resolved
Resolved ≠ Finality
Severity ≠ State
One Subject's State ≠ Whole-Harm State
Majority Improvement ≠ Collective Harm Resolution
Component Resolution ≠ Composite Harm Resolution
Temporary Non-Manifestation ≠ Resolution
System Stopped ≠ Harm Resolved
System Fixed ≠ Harm Recovered
Containment Applied ≠ Stabilized
Mitigation Applied ≠ Reduced
Recovery Program Started ≠ Recovering
Claim Closed ≠ Harm Resolved
Compensation Paid ≠ Harm Recovered
Predicted Escalation ≠ Escalating State
Simulated Recovery ≠ Recovering State
Future Risk ≠ Current Harm State
Classification Confidence ≠ Harm State
State Transition ≠ Classification Correction
- Canonical Boundary
The HSCM — Harm State Classification Module governs the classification of
an identified current Harm condition into the canonical HSS Harm States of
Emerging, Active, Escalating, Stabilized, Reduced, Recovering, Residual,
and Resolved, including material qualifications, uncertainty, confidence,
dimensional differences, subject-level differences, collective, composite,
distributed, and intermittent Harm representation, and classification
correction. It does not identify the current Harm condition, determine
whether Harm originally exists, redefine the Harm Boundary, govern State
Persistence, independently govern State Transition or Monitoring, govern
the dynamics of Harm accumulation, compounding, propagation or cascading,
perform Containment or Mitigation, govern Recovery or Residual Harm
determination, establish Resolution criteria or Finality, determine
general causation, assign accountability or liability, govern claims or
compensation, or replace operational reality, evidence, validation,
integrity, identity, AI behavior, autonomous execution, simulation,
prediction, Value, tax-event, or jurisdiction-specific governance.
- Canonical Closing Statement
A Harm State is not a label attached to whatever system surrounds the
Harm. It is a governed representation of the Harm's actual current
condition.
Stopping the mechanism does not mean the Harm is Resolved. Beginning
Recovery does not mean the Harm is Recovered. Reducing one component does
not mean a composite Harm has disappeared.
HSCM keeps those distinctions explicit.
The correct Harm State is the state supported by the Harm itself — not the
state that is most convenient to declare.