Canonical Definition

The HEBM — Harm Effect Boundary Module governs how the effects belonging
to a recognized Harm object are identified, differentiated, associated,
included, excluded, grouped, separated, expanded, contracted, disputed,
reassessed, and preserved across subjects, domains, systems, locations,
and time.

HEBM establishes the transition:

Governed Boundary Identification State + Governed Harm Subject Boundary +
Governed Harm Temporal Boundary → Effect Candidates → Effect Relationship
Assessment → Effect Inclusion / Exclusion → Governed Harm Effect Boundary

Its central question is:

Which effects legitimately belong to the governed Harm and which effects
remain outside it?

• Operational Role

A recognized Harm may produce more than one observable consequence.

It may involve:

direct effects

indirect effects

delayed effects

persistent effects

functional effects

cross-domain effects

collective effects

distributed effects

and effects that only become visible after the originating mechanism has
stopped.

HEBM determines which of those effects belong to the Harm being bounded.

Its operational role is:

to establish the legitimate effect extent of Harm without reducing Harm to
its first visible manifestation or expanding it to every remotely
associated consequence.

• Position Within HBS

The HBS internal flow is:

Boundary Identification → Subject Boundary → Temporal Boundary → Effect
Boundary → Downstream Boundary

The Core Module sequence is:

HBIM → HSBM → HTBM → HEBM → HDBM

HEBM receives:

Governed Boundary Identification State

Governed Harm Subject Boundary

Governed Harm Temporal Boundary

and produces:

Governed Harm Effect Boundary

HDBM subsequently determines how far the complete Harm Boundary extends
into downstream consequences.

• Governed Object

The primary governed object is:

Harm Effect Boundary

Possible effect-boundary states include:

Candidate Effect

Provisionally Included

Included

Partially Included

Provisionally Excluded

Excluded

Uncertain

Disputed

Reclassified

Separate Harm Candidate

Each state must remain traceable to its basis.

• Foundational Effect Principle

The foundational principle is:

The first visible effect is not necessarily the complete Harm.

Therefore:

Observed Effect ≠ Complete Harm Effect Boundary

• Effect vs Manifestation

A Harm Manifestation is how Harm becomes apparent or observable.

A Harm Effect is a materially affected condition belonging to the Harm.

The two may overlap but are not identical.

Therefore:

Manifestation ≠ Complete Harm

and:

Manifestation ≠ Effect by Default

One underlying Harm effect may generate several manifestations.

Several effects may also become visible through one manifestation.

• Effect Inclusion

An effect may be included where a materially supportable relationship
exists between the effect and the governed Harm.

The relationship may be:

direct

indirect

persistent

delayed

functional

cross-domain

collective

distributed

or:

composite

Inclusion establishes that the effect belongs within the governed Harm.

It does not establish general causation, accountability, liability, or
compensation.

• Effect Exclusion

An effect may be excluded where it is:

unrelated

insufficiently associated

duplicative

better represented as another Harm object

or:

outside the supported effect scope

Exclusion must remain reasoned and traceable.

Canonical distinction:

Excluded Effect ≠ No Harm Effect Elsewhere

• Direct Effects

A Direct Effect has a materially immediate relationship to the governed
Harm.

Examples may include:

physical impairment

system degradation

data corruption

functional loss

environmental damage

Directness may strengthen inclusion.

It does not mean the direct effect represents the complete Harm.

Therefore:

Direct Effect ≠ Complete Harm

• Indirect Effects

An effect may arise through one or more intermediate relationships.

Conceptually:

Harm → Effect A → Effect B

Effect B may still belong to the Harm where sufficient continuity exists.

Therefore:

Indirect Effect ≠ Irrelevant Effect

Indirectness requires assessment, not automatic exclusion.

• Immediate and Delayed Effects

Effects may appear:

immediately

or:

after a temporal delay

A delayed effect may belong to the same Harm even where:

Exposure has ended

the mechanism has stopped

Recognition already occurred

or:

other effects have disappeared

Canonical principle:

Delayed Appearance ≠ New Harm by Default

• Persistent Effects

An effect may continue after the originating condition has ended.

Conceptually:

Mechanism Ends → Effect Persists → Harm Continues

Therefore:

Mechanism Termination ≠ Effect Termination

and:

Effect Persistence May Extend Beyond Operational Activity

HTBM governs the temporal boundary.

HEBM governs whether the persistent effect belongs to the Harm.

• Functional Effects

Harm may alter what a subject, object, system, organization,
infrastructure, or environment can do.

Functional effects may involve:

reduced capability

loss of function

impaired performance

reduced availability

loss of continuity

restricted use

degraded operational capability

A functional effect does not require physical damage.

• Physical Damage ≠ Required Harm Effect

Canonical principle:

A Harm effect does not require physical damage in order to belong to a
governed Harm.

This preserves HAGA™ across physical and non-physical domains.

• Hidden and Latent Effects

Some effects may exist without immediate visibility.

They may be:

internal

latent

distributed

difficult to observe

difficult to measure

not yet reported

Therefore:

Invisible Effect ≠ Non-Existing Effect

HEBM must support uncertain or provisionally included hidden effects where
materially justified.

• Effect Across Multiple Subjects

One Harm effect structure may affect multiple subjects differently.

Conceptually:

Harm → Effect A on Subject 1

Harm → Effect B on Subject 2

Harm → Effect C on Subject 3

Different effects do not automatically create different Harm objects.

Therefore:

Different Subject Effects ≠ Different Harms by Default

• Multiple Effects on One Subject

One subject may experience several effects belonging to the same Harm.

Conceptually:

Harm → Physical Effect

Harm → Functional Effect

Harm → Financial Effect

The existence of multiple effects does not automatically require multiple
Harm objects.

• Effect Severity Differences

Effects within one Harm may vary significantly in severity.

Therefore:

Shared Harm ≠ Equal Effect Severity

Effect inclusion must remain separate from severity classification.

HSS later governs Harm State.

• Collective Effects

Some effects exist at:

group

organizational

infrastructure

environmental

population

or:

societal

level.

A collective effect must not be rejected merely because it cannot be
represented as identical individual effects.

Canonical principle:

Collective Effect ≠ Sum of Identical Individual Effects

• Distributed Effects

An effect may be distributed across:

systems

agents

users

locations

organizations

or:

time periods

No individual component may represent the complete effect.

HEBM must remain capable of governing the effect at the materially
coherent level.

• Composite Effects

Several individually limited consequences may collectively form a material
Harm effect.

Conceptually:

Effect Component A + Effect Component B + Effect Component C → Composite
Effect

Canonical principle:

Individual Effect Insufficiency ≠ Composite Effect Insufficiency

This is especially important for distributed AI and autonomous systems.

• Local Effect ≠ Complete Effect

A locally observed consequence may represent only part of the Harm.

Therefore:

Local Effect Boundary ≠ Complete Harm Effect Boundary

Technical decomposition must not artificially fragment one coherent effect
structure.

• Cross-Domain Effects

One Harm may produce effects across multiple domains.

Examples:

Physical → Medical / Health → Functional

Digital → Operational → Financial

Environmental → Infrastructure → Societal

Cognitive → Behavioral → Organizational

HEBM must support cross-domain effect inclusion where continuity remains
materially supportable.

• Domain Transition ≠ New Harm

Canonical distinction:

Domain Transition ≠ New Harm Object by Default

A new domain may represent another dimension of the same Harm rather than
a separate Harm lifecycle.

• Domain Boundary

A domain boundary may help classify effects.

It does not define the Harm Effect Boundary automatically.

Therefore:

Domain Boundary ≠ Effect Boundary

• System Boundary

Effects may cross:

models

agents

applications

networks

platforms

physical systems

infrastructure

Therefore:

System Boundary ≠ Effect Boundary

The effect boundary follows the Harm effect, not the system architecture.

• Organizational Boundary

Effects may cross organizational structures.

A consequence may begin in one organization and materially affect another.

Therefore:

Organizational Boundary ≠ Effect Boundary

• Location Boundary

Effects may occur across different physical or digital locations.

Therefore:

Location Change ≠ New Effect by Default

and:

Shared Location ≠ Shared Effect by Default

• Session and Agent Boundaries

AI-mediated Harm may develop across multiple sessions and agents.

Therefore:

Session Boundary ≠ Effect Boundary

Agent Boundary ≠ Effect Boundary

Tool Boundary ≠ Effect Boundary

The relevant question is whether the effect remains part of one materially
coherent Harm structure.

• Composite AI Effect

A distributed AI operation may produce an effect that cannot be attributed
to one individual output.

Conceptually:

Output₁ + Output₂ + Tool Action₃ + Agent Action₄ → Composite Harm Effect

Canonical principle:

Safe Output₁ + Safe Output₂ + Safe Output₃ ≠ Safe Composite Behavior by
Default

For HEBM:

Absence of a complete Harm effect in any single component does not
establish absence of a composite Harm effect.

• Perception-to-Action Effects

Autonomous systems may produce physical or operational effects without
passing through language.

HEBM must therefore govern effects arising through:

Perceive → Interpret → Decide → Authorize → Act

without requiring:

prompt

text output

linguistic reasoning trace

or:

explicit verbal decision

Canonical principle:

Physical contact is an operational mechanism. Harm is a governed
consequence.

• Effect Relationship Structure

HEBM should preserve material relationships such as:

Harm → Effect

Effect → Manifestation

Effect A → Effect B

Effect → Subject

Effect → Domain

Effect → System

Effect → Time

Effect → Downstream Candidate

Relationship mapping supports effect-boundary governance.

It does not establish general causation.

• Effect Relationship ≠ General Causation

Canonical distinction:

Effect Relationship ≠ Final Causal Attribution

HAGA™ may require sufficient causal association for Harm governance.

It does not become a general forensic or causation architecture.

• Effect Chain

A Harm may produce an Effect Chain:

Primary Effect → Secondary Effect → Tertiary Effect

HEBM determines which effects remain part of the Harm effect structure.

HDBM later determines how far downstream the complete Harm Boundary
legitimately extends.

• Effect vs Downstream Consequence

HEBM and HDBM must remain distinct.

HEBM asks:

Which effects belong to the recognized Harm?

HDBM asks:

How far should the Harm Boundary extend through downstream consequences
and relationships?

Therefore:

Effect Inclusion ≠ Unlimited Downstream Inclusion

• Intervening Conditions

An effect relationship may contain an intervening condition.

Conceptually:

Harm → Effect A → Independent Condition → Effect B

The independent condition may indicate:

Boundary Break

or:

Separate Harm Candidate

HEBM must preserve the distinction for downstream governance.

• One Harm vs Multiple Harms

Several effects may represent:

one Harm with multiple effects

or:

multiple distinct Harm objects

HEBM must support both possibilities.

Relevant factors include:

effect continuity

subject relationships

temporal continuity

domain structure

independent conditions

separate Harm Emergence

No single factor should determine the answer automatically.

• Effect Merge

Effects initially treated separately may later be established as
manifestations or components of one coherent Harm effect structure.

HEBM must support:

Effect Structure A + Effect Structure B → Unified Effect Boundary

without erasing prior states.

• Effect Separation

An effect initially treated as part of one Harm may later be established
as:

unrelated

or:

part of another Harm object

HEBM must support separation while preserving the previous inclusion state
and reason for change.

• Duplicate Effect Control

One effect may be represented through multiple:

reports

systems

measurements

claims

records

descriptions

Multiple representations must not create multiple effects automatically.

Canonical principle:

Information Duplication ≠ Effect Multiplication

• Effect Uncertainty

Where effect inclusion cannot yet be established, HEBM may use:

Provisionally Included

Provisionally Excluded

Uncertain

Disputed

rather than forcing a binary conclusion.

Canonical principle:

Uncertain Effect Boundary ≠ No Effect

• Effect Confidence

Effect-boundary confidence may vary across different effects.

Example:

Direct Physical Effect — High

Functional Effect — High

Delayed Secondary Effect — Moderate

Cross-Domain Effect — Low / Pending

A single confidence score must not erase meaningful distinctions.

• Effect Confidence ≠ Severity

Canonical distinction:

Effect Boundary Confidence ≠ Harm Severity

A severe effect may be difficult to bound.

A minor effect may be precisely established.

• Effect Boundary Inflation

Effect Boundary Inflation occurs where insufficiently associated
consequences are incorporated into the Harm.

Potential causes include:

temporal proximity alone

shared location alone

shared subject alone

shared system alone

correlation alone

financial consequence alone

remote association

HEBM must prevent unsupported expansion.

• Effect Boundary Suppression

Effect Boundary Suppression occurs where materially associated effects are
excluded because they are:

indirect

delayed

hidden

cross-domain

distributed

collective

outside the original system

outside the original organization

not included in the initial report

HEBM must prevent artificial narrowing.

• Visibility Bias

Highly visible effects may dominate governance while less visible effects
remain ignored.

Canonical principle:

Effect Visibility ≠ Effect Legitimacy

• Reporting Bias

A reported effect may be easier to govern than an unreported effect.

But:

Reported Effect ≠ More Real Effect

and:

Unreported Effect ≠ Boundary Irrelevance

• Evidence Availability Bias

Evidence may be stronger for some effects than others.

Therefore:

Evidence Availability ≠ Effect Reality

Insufficient evidence may require uncertainty rather than automatic
exclusion.

• Claim and Liability Bias

Claims and liability structures may emphasize particular effects.

HEBM must remain neutral.

Therefore:

Claimed Effect ≠ Harm Effect by Default

Unclaimed Effect ≠ Non-Harm Effect

Liability-Relevant Effect ≠ Complete Harm Effect Boundary

• Value Consequence Boundary

A Harm effect may create Value consequences.

VFM® governs what happens to Value.

HEBM must not equate Value change with Harm effect automatically.

Therefore:

Value Loss ≠ Harm Effect by Default

and:

No Measurable Value Loss ≠ No Harm Effect

• Effect Persistence and Recovery

An effect may:

persist

reduce

stabilize

recover

or:

leave Residual Harm

HEBM establishes whether the effect belongs to the Harm.

Later HAGA™ standards govern its State, Mitigation, Recovery, and Residual
Harm.

Therefore:

Effect Inclusion ≠ Effect State Governance

• Effect Boundary Expansion

The Effect Boundary may expand where new materially associated effects
become known.

Conceptually:

Effect Boundary₁ → New Supported Effect → Effect Boundary₂

Expansion must preserve the previous boundary.

• Effect Boundary Contraction

The Effect Boundary may contract where a previously included effect is
later shown not to belong.

Contraction must preserve:

previous inclusion

new basis

reason for exclusion

effective change

• Effect Reclassification

An effect may be reclassified from:

direct → indirect

individual → collective

single-domain → cross-domain

effect → manifestation

effect → separate Harm candidate

without erasing its history.

• Effect Boundary Reopening

A previously stable Effect Boundary may require reopening where:

delayed effects appear

hidden effects become observable

cross-domain consequences emerge

distributed effects are correlated

new subject information appears

historical information changes the effect structure

• Historical Effect Boundary Integrity

HEBM must preserve:

Effect Boundary State₁ → New Information → Effect Boundary State₂

including:

effect

classification

inclusion state

relationship basis

confidence

uncertainty

dispute

change reason

Canonical principle:

Current Effect Boundary ≠ Complete Effect Boundary History

• Relationship to Adjacent HBS Modules

HBIM asks:

Which effects may be boundary-relevant?

HSBM establishes:

Who or what belongs to the Harm?

HTBM establishes:

What period belongs to the Harm?

HEBM asks:

Which effects belong to that Harm?

HDBM next asks:

How far downstream does the Harm Boundary legitimately extend?

The modules are connected but must not collapse into one another.

• Cross-Architecture Boundaries

ORA™ governs the operational reality through which effects may arise.

AIG® governs AI behavior.

ASGA™ governs autonomous execution.

ADIT® governs evidence.

VALIDOS™ governs validation.

INTEGROS® governs integrity.

OBIDENITY® governs identity and provenance.

VFM® governs Value lifecycle.

AGA™ governs accountability.

HEBM uses relevant outputs from these spaces where necessary but governs
only:

which effects belong to the Harm object.

• Minimum Implementation Framework

A minimum HEBM implementation requires eight capabilities.

• Effect Identification & Classification

Represent direct, indirect, immediate, delayed, persistent, functional,
hidden, collective, distributed, composite, and cross-domain effects.

• Effect Inclusion & Exclusion

Apply traceable states for Included, Provisionally Included, Excluded,
Provisionally Excluded, Uncertain, Disputed, and Separate Harm Candidate
effects.

• Effect Relationship Governance

Represent effect-to-Harm, effect-to-subject, effect-to-effect,
manifestation, temporal, domain, system, and downstream relationships
without converting association into general causation.

• Composite & Distributed Effect Governance

Recognize materially coherent effects formed across multiple actions,
agents, systems, sessions, subjects, locations, or time periods.

• Cross-Domain & Persistence Governance

Preserve effects that cross domains or persist beyond Exposure, mechanism
operation, Recognition, or initial manifestation.

• Boundary Bias Controls

Prevent visibility, reporting, technical systems, organizations, claims,
liability, Value consequences, or evidence availability from automatically
defining effect inclusion.

• Reassessment & Historical Continuity

Support effect-boundary expansion, contraction, reclassification, dispute,
reopening, merge, separation, and correction while preserving prior
states.

• Governed Output

Produce a traceable Governed Harm Effect Boundary suitable for integration
with the Downstream Boundary and complete Governed Harm Boundary.

A system that treats only the first visible, directly measurable, or
administratively recorded effect as Harm does not fully implement HEBM.

• Use Case 1 — One Physical Harm Produces Effects Across Multiple

Domains

A physical Harm initially becomes visible through an immediate physical
effect.

Later, the same Harm produces:

persistent functional impairment

and:

a materially related financial consequence

HEBM assesses whether these effects remain part of the same Harm rather
than assuming:

only the physical effect counts

or:

every financial consequence automatically belongs

The supported effects are incorporated into the Governed Harm Effect
Boundary, while insufficiently related consequences remain outside or
uncertain.

The governing principle is:

The first manifestation may reveal Harm, but it does not necessarily
define the complete effect structure of that Harm.

• Use Case 2 — Distributed AI Produces a Composite Effect

An AI operation is distributed across several models, agents, sessions,
and tools.

No individual output creates the complete harmful effect.

When combined, however, the operation produces a materially coherent
consequence affecting the governed subject.

HEBM evaluates the effect at the composite level rather than limiting
analysis to individual outputs.

The resulting Harm Effect Boundary may therefore cross:

agent

session

tool

and:

system

boundaries.

The governing principle is:

An effect produced by the whole must remain governable even when no
individual component contains the whole effect.

• Governance Requirements

HEBM requires that:

• the first visible effect remain distinct from the complete Harm
  Effect

Boundary; - manifestations remain distinct from Harm and from effects
where necessary; - direct and indirect effects remain governable; -
delayed appearance not automatically create a new Harm; - persistent
effects remain governable after mechanism termination; - physical damage
not be required for Harm Effect inclusion; - hidden and latent effects
remain representable; - multiple effects on one subject not automatically
create multiple Harms; - different effects across subjects not
automatically create separate Harms; - collective, distributed, and
composite effects remain governable; - individual insufficiency not
automatically establish composite insufficiency; - domain, system,
organization, location, session, agent, and tool boundaries not
automatically define Effect Boundary; - non-linguistic autonomous effects
remain governable; - effect relationships remain distinct from final
causal attribution; - Effect Boundary remain distinct from unlimited
Downstream Boundary; - intervening independent conditions remain capable
of supporting Boundary Break; - effect merge, separation, duplication,
uncertainty, and dispute remain representable; - visibility, reporting,
evidence availability, claim structure, liability, and Value consequences
not determine Effect Boundary by default; - effect inclusion remain
distinct from Harm State and Recovery governance; - expansion,
contraction, reclassification, reopening, and correction preserve
historical integrity.

• Governance Logic

HEBM applies the following logic:

Have Effect Candidates been identified?

If no:

→ maintain the Effect Boundary as incomplete.

If yes:

→ classify candidate effects.

Is the effect materially associated with the recognized Harm?

If sufficiently supported:

→ include it.

If plausible but unresolved:

→ preserve Provisional or Uncertain state.

If unsupported:

→ exclude it.

Is the effect indirect, delayed, hidden, or cross-domain?

→ do not exclude it solely for that reason.

Does the effect cross technical or organizational boundaries?

→ follow the effect rather than the system boundary.

Do several individually insufficient effects form one material composite
effect?

If yes:

→ assess the composite structure.

Is the effect merely another representation of an existing effect?

→ prevent duplication.

Does an independent condition separate the effect from the original Harm?

If yes:

→ identify potential Boundary Break or Separate Harm Candidate.

Is the effect being included only because it is temporally close,
financially relevant, claimed, or highly visible?

→ reassess the material relationship.

Has new information changed the effect structure?

→ expand, contract, reclassify, separate, merge, or reopen while
preserving history.

At every stage:

Manifestation ≠ Complete Harm

Direct ≠ Complete

Indirect ≠ Irrelevant

Delayed ≠ Separate Harm

System Boundary ≠ Effect Boundary

Correlation ≠ Inclusion

Effect Inclusion ≠ General Causation

Effect Inclusion ≠ Liability

• Governance Outputs

HEBM may produce:

• Governed Harm Effect Boundary
• Included Effect Set
• Provisionally Included Effect Set
• Excluded Effect Set
• Uncertain Effect Set
• Disputed Effect Set
• Direct Effect State
• Indirect Effect State
• Delayed Effect State
• Persistent Effect State
• Functional Effect State
• Hidden / Latent Effect State
• Collective Effect Structure
• Distributed Effect Structure
• Composite Effect Structure
• Cross-Domain Effect Structure
• Effect Relationship Map
• Effect Chain
• Potential Boundary Break
• Separate Harm Candidate
• Duplicate Effect Resolution
• Effect Boundary Confidence
• Effect Boundary Expansion Record
• Effect Boundary Contraction Record
• Effect Reclassification Record
• Effect Boundary Reopening State
• Historical Effect Boundary Record

The principal output is:

Governed Harm Effect Boundary

• Critical Distinctions

Observed Effect ≠ Complete Harm Effect Boundary

Manifestation ≠ Complete Harm

Manifestation ≠ Effect by Default

Excluded Effect ≠ No Harm Effect Elsewhere

Direct Effect ≠ Complete Harm

Indirect Effect ≠ Irrelevant Effect

Delayed Appearance ≠ New Harm by Default

Mechanism Termination ≠ Effect Termination

Physical Damage ≠ Required Harm Effect

Invisible Effect ≠ Non-Existing Effect

Different Subject Effects ≠ Different Harms by Default

Shared Harm ≠ Equal Effect Severity

Collective Effect ≠ Sum of Identical Individual Effects

Individual Effect Insufficiency ≠ Composite Effect Insufficiency

Local Effect Boundary ≠ Complete Harm Effect Boundary

Domain Transition ≠ New Harm Object by Default

Domain Boundary ≠ Effect Boundary

System Boundary ≠ Effect Boundary

Organizational Boundary ≠ Effect Boundary

Session Boundary ≠ Effect Boundary

Agent Boundary ≠ Effect Boundary

Tool Boundary ≠ Effect Boundary

Effect Relationship ≠ Final Causal Attribution

Effect Inclusion ≠ Unlimited Downstream Inclusion

Information Duplication ≠ Effect Multiplication

Uncertain Effect Boundary ≠ No Effect

Effect Boundary Confidence ≠ Harm Severity

Effect Visibility ≠ Effect Legitimacy

Evidence Availability ≠ Effect Reality

Claimed Effect ≠ Harm Effect by Default

Unclaimed Effect ≠ Non-Harm Effect

Value Loss ≠ Harm Effect by Default

No Measurable Value Loss ≠ No Harm Effect

Effect Inclusion ≠ Effect State Governance

• Canonical Boundary

The HEBM — Harm Effect Boundary Module governs which direct, indirect,
immediate, delayed, persistent, functional, hidden, collective,
distributed, composite, and cross-domain effects legitimately belong
within the effect boundary of a recognized Harm object. It governs effect
classification, inclusion, exclusion, relationships, manifestations,
persistence, composite formation, cross-domain continuity, uncertainty,
dispute, duplication, expansion, contraction, reclassification,
separation, reopening, and historical Effect Boundary continuity. It does
not establish the complete Downstream Boundary, formally determine Harm,
establish general causation, assign accountability or responsibility,
determine liability or compensation, govern Harm State, propagation,
containment, mitigation, recovery, or replace operational reality,
evidence, validation, integrity, identity, AI behavior, autonomous
execution, simulation, Value, or tax-event governance.

• Canonical Closing Statement

Harm is rarely exhausted by the first effect through which it becomes
visible.

Its effects may be direct or indirect, immediate or delayed, visible or
hidden, individual or collective, local or distributed, and may cross
domains and systems without ceasing to belong to the same Harm.

HEBM ensures that those effects are neither excluded because they are
inconveniently distant from the first manifestation nor included merely
because some association can be found.

The Effect Boundary follows the materially supportable consequences of
Harm — not the limits of visibility, reporting, technology, or
administrative convenience.