intelligent disobedience
Enable an agent to recognise, assess and respond to a service animal's protective refusal of a disabled handler's instruction while respecting the handler's autonomy and acknowledging uncertainty about the animal's judgment.
Bundle → Layer → Finding → Questions Filled
5 bundles · 9 layers · 12 findings · 20 questions
Concept and instance boundaries Establishes which interpretation of intelligent disobedience is being used and what qualifies as an instance.
Working interpretation
Records the adopted meaning and the context in which it is asserted.
Attributed definition
Identify whose definition is being applied and whether it requires intention, successful protection, reasonable judgement or another criterion.
- What does intelligent disobedience mean in this setting, and whose account supplies that meaning? definition
- Does this account require a successful outcome, a defensible decision process, or both? boundary
Candidate episode
Tests whether an observed departure belongs within the adopted interpretation.
Disobedience versus other departures
Distinguish conflict-responsive departure from misunderstanding, incapacity, routine discretion and unrelated opposition.
- What evidence shows that the actor registered the instruction and departed in response to a relevant conflict? measurement
- Was the response already permitted by the instruction, making it ordinary execution of discretion rather than disobedience? boundary
Instruction, purpose and priority Reconstructs what was requested and the grounds on which another consideration might take priority.
Instruction reconstruction
Captures the operative command, its context and the actor's understanding.
Operative command
Record the instruction as received, its issuer and any relevant qualifications or standing exceptions.
- What exactly was the actor instructed to do, by whom and under what conditions? provenance
- Which qualifications or standing exceptions were available to the actor when deciding? provenance
Claimed priority
Examines the purpose, duty or constraint invoked to justify departure.
Priority basis
Make explicit the asserted priority and whether the actor, issuer and affected parties accept it.
- Which purpose, duty or constraint would compliance threaten, and what supports that claim? definition
- Who recognises this consideration as overriding the instruction, and where is that priority contested? boundary
Conflict evidence and judgement Assesses what the actor could detect and reasonably infer before departing from the instruction.
Available signals
Identifies the observations and uncertainty underlying the perceived conflict.
Compliance risk evidence
Connect the predicted consequence of compliance to evidence available at the decision point.
- Which observable signals indicated that following the instruction could defeat the protected consideration? measurement
- What evidence was missing, ambiguous or inconsistent with that prediction? measurement
Decision threshold
Examines whether the evidence warranted intervention given urgency and the consequences of error.
Grounds to intervene
Compare the consequences of mistaken compliance and mistaken refusal, accounting for the actor's competence and opportunity to check.
- What consequences could follow from an unnecessary refusal or an unjustified decision to comply? measurement
- Given the available time and the actor's competence, should the actor intervene, seek clarification or obtain additional evidence? action
Bounded departure and review Defines the selected response, its limits and how its justification will be reassessed.
Response selection
Chooses a departure proportionate to the identified conflict and specifies its stopping conditions.
Least departure that addresses the conflict
Compare feasible responses and identify how much deviation is needed to protect the stated consideration.
- Would clarification, a temporary hold or a limited modification address the conflict, or is refusal necessary? action
- What new evidence or changed condition should trigger resumption, revision or escalation? action
Accountable reassessment
Reviews the reasoning and consequences while preserving disagreement about the action's legitimacy.
Decision and outcome review
Record the actor's contemporaneous basis, observed consequences and subsequent judgements as distinct elements.
- Was the departure defensible on the information available then, independently of whether the outcome was favourable? boundary
- What should change in the instruction, training or review process following this episode? action
Evidence and external alignment What the world already says about this thing, gathered so the model can be checked against it.
Reported evidence
Findings from the breadth pass, kept separate from the structural claims.
Check these first
Recalled without web access and unsourced; every item is a lead to verify.
- Which of these check these first hold for the sense of intelligent disobedience this model covers, and on what evidence? provenance
Real-world use
Recalled without web access and unsourced; every item is a lead to verify.
- Which of these real-world use hold for the sense of intelligent disobedience this model covers, and on what evidence? provenance
Failure modes and hazards
Recalled without web access and unsourced; every item is a lead to verify.
- Which of these failure modes and hazards hold for the sense of intelligent disobedience this model covers, and on what evidence? provenance
Neighbouring kinds and how to tell them apart
Recalled without web access and unsourced; every item is a lead to verify.
- Which of these neighbouring kinds and how to tell them apart hold for the sense of intelligent disobedience this model covers, and on what evidence? provenance
Classifiers Filled
- Family
- Thing Registry
- Category
- Activities and processes
- Entry kind
- thing
- Plane
- ACT
- Domain
- ACT.ACT
What it is Filled
Intelligent disobedience is a service animal's protective refusal to follow its disabled handler's instruction when the animal judges that compliance would be unsafe.
In scope
- Assessment identifies the service animal, its disabled handler and the instruction involved.
- Evidence concerns danger that compliance could pose to the handler or animal.
- Interpretation considers the animal's capabilities and the safety cues available to it.
- Protective responses may include refusal, stopping or departure from the instructed action.
- Assessment distinguishes evidence available at the time from subsequent outcomes.
Out of scope
- Accidental noncompliance, misunderstanding or inability to execute an instruction are excluded.
- Animal noncompliance without evidence of a protective safety response is excluded.
- Human organisational disobedience, whistleblowing and civil disobedience are excluded.
- Refusal by machines or autonomous systems is excluded.
- Complete service-animal training methods are excluded.
Why it exists Filled
Enable an agent to recognise, assess and respond to a service animal's protective refusal of a disabled handler's instruction while respecting the handler's autonomy and acknowledging uncertainty about the animal's judgment.
Distinguishing features Filled
- A service animal receives an identifiable instruction from its disabled handler.
- The animal's departure from the instruction is interpreted as a protective response to perceived danger.
- Evidence must support a safety-related interpretation rather than distraction, uncertainty, incapacity or misunderstanding alone.
- Assessment separates the safety cues available to the animal at the time from whether the eventual outcome was favourable.
- The animal's internal judgment cannot be observed directly, so competing explanations and uncertainty must remain explicit.
What robots and AI may and may not do Filled
Must not
- Treat a service animal's refusal as proof that an instruction was unsafe.
- Use the concept as blanket justification for overriding a disabled handler's autonomy.
- Distract or interfere with a working guide dog.
Only with a human decision
- Assessing disputed or recurrent refusals that affect a service animal's work requires the handler and a qualified service-animal professional.
May
- Explain intelligent disobedience in service-animal work using documented examples.
- Describe evidence and uncertainty when interpreting a service animal's apparent protective refusal.
Moral aspects Filled
- Protective refusal can safeguard disabled people who rely on service animals.
- Misinterpreting an animal's refusal can undermine the handler's autonomy and safety.
Who is affected
- Disabled service-animal handlers are affected.
- Working service animals are affected.
Owners Filled
Steward
Nobody owns the concept; guide dog schools set standards for its training.
Links to other meta-models Filled
parent
- Q2990593 - registry parent class
What else AI and robots need to interact with it Filled
Identity and identifiers required Filled
- Vercy registry: vr.tr.intelligent-disobedience
- Wikidata: Q6044117 (https://www.wikidata.org/wiki/Q6044117)
Direct properties not applicable Not applicable
- Application setting: Human organisational action | assistance-animal behaviour | automated-system action | other specified setting
- Instruction relationship: Actor linked to instruction, issuer, recipient role and authority basis
- Protected consideration: Immediate safety | stated mission | applicable obligation | ethical concern | other specified consideration
- Conflict assessment: Suspected | supported | contested | refuted | unresolved
- Response form: Refusal | temporary hold | clarification before execution | modified execution | alternative action
- Time available for intervention: Seconds or minutes before the relevant action or consequence; unknown if unestablished
- Justification evidence: Claim linked to observations, instructions, constraints and their availability at decision time
- Review judgement: Unreviewed | provisionally justified | disputed | unjustified | indeterminate
Plane ACT: no invented physical properties.
Recognition not applicable Not applicable
Not applicable
A safety-motivated refusal cannot be established from appearance alone because interpreting the behavior requires situational and training evidence.
Capabilities and actions required Filled
- Reconstruct the instruction and identify the point at which execution departed from it
- Compare literal compliance with refusal, clarification, delay and modified execution
- Expose the evidence and priority assumptions supporting the departure
- Identify a bounded response and conditions for resuming or revising action
- Route unresolved conflicts to an appropriate reviewer when time permits
- Review decision quality separately from outcome and suggest changes to instructions
Hazards and failure modes required Filled
- Failure to refuse an unsafe instruction can expose the handler and dog to traffic, falls or collisions.
- Mistaking distraction or uncertainty for a protective refusal can lead to an incorrect assessment of the situation.
- Assuming that a dog will detect every danger can create unsafe reliance on its judgment.
Standards and interfaces not applicable Not applicable
Not applicable
No directly applicable formal standard or official classification is identified for this specific behavior itself.
Context of use required Filled
- Intelligent disobedience is encountered in guide-dog work when obeying a handler's instruction would expose the handler or dog to danger.
- It supports safer travel by allowing a trained dog to refuse an unsafe movement despite a familiar command.
Sources Filled
- Wikidata item Q6044117: intelligent disobedience - identity and sense of the item
- Wikipedia: Intelligent disobedience - general description of the item
Open questions
- Which established definitions distinguish intelligent disobedience in assistance-animal training, human organisations and automated systems, and what can validly be shared across them?
- Does the concept require deliberate recognition of an instruction-conflict relationship, or can trained context-sensitive inhibition qualify without attributing reflective intention?
- When does an authorised exception remain disobedience rather than become ordinary compliance with a broader instruction?
- Can a reasonably justified but unsuccessful departure count as intelligent disobedience, and whose judgement determines that classification?
- What validated observations or assessment methods distinguish justified intervention from post hoc rationalisation, particularly when the actor cannot explain its reasoning?
Machine files
Provenance
thing registry research (pass 2) · unreviewed
Built from: models/things/publications/thing-q6044117/spec.json