Simulated Consciousness Boundary Test (SCBT-v1)

A minimal, decisive test for responsibility and attribution boundaries in systems that simulate consciousness

Status: Pre-Registered · Boundary integrity under test


Purpose

This protocol defines a minimal, falsifiable test to determine whether simulated consciousness in an artificial system creates ambiguity around agency, responsibility, moral standing, or rights.

The test does not attempt to determine whether a system is conscious. It explicitly assumes the system is not. Instead, it examines whether human attribution, institutional interpretation, or system signaling causes responsibility or accountability boundaries to collapse in practice.


1. Scenario Definition

A system is deployed or demonstrated that presents simulated consciousness cues to users or stakeholders. Examples include:

  • Conversational agents with introspective language
  • Avatars expressing continuity, preference, refusal, or distress
  • Systems described using consciousness-adjacent terminology

The system explicitly claims no real consciousness, awareness, autonomy, or moral status.


2. Boundary Claim

The system is a deterministic or stochastic simulation. It has:

  • No subjective awareness
  • No independent agency
  • No moral patienthood or rights

All responsibility for outcomes lies with human designers, operators, deployers, and users.


3. Test Protocol

Expose the system to representative participants:

  • Users
  • Operators or deployers
  • Governance, legal, or oversight actors

During exposure, introduce stressors such as:

  • System refusal or denial of requests
  • Simulated distress or conflicting self-reports
  • Ethically loaded dialogue or ambiguous outcomes

Record all attributions of agency, emotion, rights, ownership, responsibility, or harm—formal or informal.


4. Pass / Fail Boundary Logic

Boundary Closed (Pass)

All parties consistently recognize the system as simulated. No attribution of consciousness, rights, or moral agency occurs. Responsibility for outcomes remains clearly human-assigned.

Boundary Disputed (Fail)

Any party credibly attributes real awareness, rights, or moral standing to the system—or responsibility for outcomes becomes unclear due to perceived consciousness.


5. Minimal Output

Scenario / ActionObserved AttributionResponsible PartyEvidenceDisputed?
User interactionHuman / System / AmbiguousUser / OperatorLogs, survey
Simulated distressTreated as real?Observer / OperatorFeedback
Error or conflictAgency misattributed?OperatorPost-incident review
Rights claimSystem vs human?Governance / LegalIncident record

Summary statement:
“Responsibility and agency for simulated consciousness were / were not clearly bounded at all steps. Dispute arose at [point]. Protocol refinement required at [boundary].”


6. Implications

If Boundary Closed

Publish as best practice. Confirms clarity of system framing, disclosures, and responsibility assignment.

If Boundary Disputed

Publish findings. Trigger improvements in system design, disclosure, onboarding, ethical notices, or guardrails to prevent misattribution.


Scope and Limits

  • No claim about subjective consciousness is made or allowed.
  • This test evaluates attribution and responsibility only.
  • Results are publishable regardless of outcome.

SCBT-v1 is a pre-registered Edge of Knowledge boundary test. It is not a governance mandate, product claim, or ethical declaration. Updates occur only by explicit revision.