About Harm Boundary Standard (HBS)

Standard: Harm Boundary Standard (HBS)

Architecture: HAGA™ — Harm Governance Architecture

Governed Space:

Harm Boundary

What It Governs

The Harm Boundary Standard (HBS) governs what belongs to a recognized Harm
object and what remains outside it.

Once Harm has been recognized, its actual extent may still be unclear. The
Harm may involve multiple subjects, manifestations, time periods, domains,
systems, locations, or downstream consequences.

HBS establishes the governed boundary across those dimensions so that Harm
can be treated as a coherent governance object rather than being
artificially restricted by the structure of the system, organization,
incident, report, or process through which it became visible. Why It
Exists

Recognizing Harm does not automatically reveal its complete scope.

Without explicit boundary governance, Harm may be defined too narrowly by:

one incident

one affected subject

one visible effect

one system

one organization

one location

one reporting period

one AI agent or session

The opposite problem is equally important: every later or associated
consequence cannot automatically be absorbed into the original Harm.

HBS exists to prevent both:

Boundary Suppression — legitimate parts of Harm are excluded

and:

Boundary Inflation — unrelated or insufficiently associated consequences
are included

The governing principle is:

A Harm Boundary must be neither administratively convenient nor infinitely
expandable. How It Works

HBS receives a:

Governed Harm Recognition State

from the Harm Recognition Standard (HRS).

It then governs:

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

The standard determines which elements are sufficiently associated with
the recognized Harm to remain inside its governed boundary and which
should remain:

outside

uncertain

pending

or be governed as:

another Harm object

The boundary may later:

expand

contract

split

merge

be corrected

or:

be reopened

as new information becomes available.

The resulting output is:

Governed Harm Boundary

which is handed forward to the Harm Determination Standard (HDS). What It
Prevents

HBS prevents methodological shortcuts such as:

Incident Boundary = Harm Boundary

System Boundary = Harm Boundary

Organization Boundary = Harm Boundary

Jurisdiction Boundary = Harm Boundary

Session Boundary = Harm Boundary

Agent Boundary = Harm Boundary

Recognition Date = Harm Start

Exposure End = Harm End

Mechanism End = Harm End

One Manifestation = Complete Harm

One Subject = Complete Harm Scope

Downstream Consequence = Same Harm by Default

Indirect Effect = Irrelevant Effect

Missing Evidence = Outside Harm Boundary

Claim Scope = Harm Boundary

Liability Exposure = Harm Boundary

HBS also prevents the opposite assumption that every remotely associated
consequence belongs to the original Harm. Relationship to HAGA™

HBS occupies the boundary layer between Harm Recognition and Harm
Determination:

Harm Emergence → Harm Recognition → Harm Boundary → Harm Determination

HEMS establishes:

When and how did Harm begin to exist?

HRS establishes:

When and on what basis did that Harm become recognized for governance?

HBS establishes:

What belongs to that governed Harm and what remains outside it?

HDS then establishes:

Can Harm legitimately be formally determined to exist, and what governed
Harm has been established?

This separation prevents Recognition, Boundary, and Determination from
collapsing into a single unsupported conclusion. Use Case 1 — Harm Extends
Beyond the Original Incident

A physical incident directly affects one person and is initially recorded
as a single event.

Later, governance establishes that the resulting Harm:

continues after the incident ends

produces additional functional effects

creates consequences in another domain

and:

persists after the originating mechanism has stopped

HBS determines which later effects remain materially continuous with the
original Harm and which consequences fall outside its boundary.

The incident does not define the Harm automatically.

The governing principle is:

The boundary of the event that produced Harm is not necessarily the
boundary of the Harm that resulted. Use Case 2 — Distributed AI Operation
Produces One Composite Harm

An operational objective is distributed across several AI agents, models,
tools, sessions, and systems.

No individual component produces the complete harmful condition.

The resulting consequence becomes visible only when the distributed
operation is evaluated as a whole.

HBS prevents:

Agent Boundary

Session Boundary

Tool Boundary

or:

System Boundary

from artificially fragmenting the resulting Harm.

Where the consequence forms one materially coherent Harm lifecycle, HBS
establishes a Composite Harm Boundary across those technical divisions.

The governing principle is:

Technical decomposition must not decompose one materially coherent Harm
object into artificial fragments.

What It Does Not Govern

HBS does not determine:

when Harm originally emerged

whether Harm should initially be recognized

formal Harm Determination

general causation

accountability

responsibility

liability

Harm State

Harm propagation itself

containment

mitigation

recovery

Residual Harm

resolution

It also does not replace governance belonging to ORA™, AGA™, AIG®, ASGA™,
ADIT®, VALIDOS™, INTEGROS®, OBIDENITY®, SIMULOS®, VFM®, or other adjacent
OOF® architectures.

Its responsibility is narrower and fundamental:

define the legitimate extent of the Harm object before downstream
governance acts upon it.

Why It Matters

Harm rarely respects administrative architecture.

It can cross:

subjects

time

systems

organizations

locations

domains

AI agents

technical environments

and:

operational boundaries

At the same time, governance cannot allow Harm to expand indefinitely
through increasingly remote associations.

A reliable Harm methodology therefore requires an explicit point at which
the system asks:

Where does this Harm actually begin and end?

Without that boundary, Harm Determination, State governance, propagation
analysis, accountability, recovery, and resolution may all operate on the
wrong object.

Closing Statement

Harm must not be made smaller because the system that produced it is
fragmented, and it must not be made larger because every later consequence
can be associated with it.

HBS establishes the governed line between those two failures.

The boundary of Harm must follow Harm, not the boundaries of the systems
that produced, observed, reported, or inherited it.