← Back to catalogue
Research draft

Three Laws of Robotics

vr.tr.three-laws-of-robotics · INF.KNW

Enable an agent to identify the Three Laws of Robotics, distinguish their formulations, and assess proposed interpretations or applications against their ordered obligations and unresolved ambiguities.

Thing Registry Information and virtual systems

Research draft, second pass

A second pass drafted this model: the structure a model of this thing needs, and what is known about it in the world. The line under this one says how the second half was obtained - researched against sources, or recalled without web access, in which case nothing here was read anywhere and every claim is a lead to verify. Unreviewed either way.

recalled by Codex without web access - no source was read

Researched by: Codex

Purpose and description

Enable an agent to identify the Three Laws of Robotics, distinguish their formulations, and assess proposed interpretations or applications against their ordered obligations and unresolved ambiguities.

The Three Laws of Robotics are Isaac Asimov's fictional, hierarchically ordered constraints on robot behaviour, prioritising prevention of harm to humans over obedience to human orders and robot self-preservation.

It can be Compare a purported statement of the laws with an identified textual formulation.; Trace how higher-priority obligations constrain a proposed command or self-protective action.; Construct scenarios that expose ambiguity about harm, inaction, or human identity.; Separate narrative outcomes from conclusions inferred under additional assumptions.; Link adaptations and extensions while preserving the identity of the three-law system.; Assess which additional definitions and evidence an operational application would require..

Distinguishing features

The referent is Asimov's fictional behavioral rule system, rather than a physical law or a robotics discipline.

Human protection includes harm permitted through inaction, so a prohibition on direct injury alone is an incomplete formulation.

Obedience is conditional on human protection, and self-preservation is subordinate to both; three equally weighted objectives do not preserve this structure.

A formulation adding an obligation toward humanity above individual humans must be identified as a Zeroth-Law extension rather than silently treated as the original three.

Textual identification is supported by the law statements and extension discussion in the [Isaac Asimov FAQ mirror](https://stason.org/TULARC/education-books/isaac-asimov/4-13-What-are-the-Laws-of-Robotics-anyway.html).

Scope

+ The three obligations concerning human protection, obedience to humans, and robot self-preservation

+ Priority and exception relations among the laws

+ Interpretation of human, harm, inaction, order, and continued existence

+ Textual provenance and differences between formulations

+ Scenario analysis and limits of claims made from the laws

- Robotics as an engineering discipline

- Robot hardware, control architectures, and implementation specifications

- Asimov's biography and complete bibliography

- Complete models of individual stories, fictional robots, or adaptations

- Legal requirements, safety certification, and general AI ethics frameworks

- The Zeroth Law as an independently developed doctrine beyond its relationship to the original three

Characteristics

Formulation relationship
Three-law formulation; paraphrase; translation; modified formulation; Zeroth-Law extension; unresolved Prevents wording changes or additional obligations from being mistaken for equivalent statements.
Obligation precedence
Human protection precedes obedience; both precede robot self-preservation; record deviations explicitly Determines which obligation constrains another when they conflict.
Inaction coverage
Explicitly included; omitted; qualified; ambiguous Separates avoiding injury from an obligation to prevent harm.
Interpretation of harm
Physical; psychological; mixed; other explicitly defined interpretation; unspecified Scenario conclusions depend on what consequences count as harm.
Protected-human interpretation
Explicit membership criterion; narrative-dependent criterion; disputed; unspecified Records whose protection and commands activate the laws without inventing a universal recognition rule.
Textual witness
Work, edition, passage locator, language, and relationship to the formulation under review Makes claims about wording and narrative interpretation traceable.
Scenario assessment state
Consistent under stated assumptions; inconsistent under stated assumptions; underdetermined; no feasible option established Supports qualified judgments without treating the laws as a complete decision procedure.

Where this came from

wikidata · CC0 1.0

Drafted structure

Bundle to layer to finding to question, as the second pass will find it: 6 bundles · 11 layers · 14 findings · 24 questions.

Textual identity Establish which statement of the Three Laws is being modeled and where it is attested.

Small textual changes can alter the rule system, so an agent needs an identifiable formulation before interpreting it.

Formulation witnesses

Locate the wording in an identifiable work and edition.

Attested law statements

Record the passage supporting each obligation, distinguishing quotation, translation, and paraphrase.

  1. Which work, edition, and passage attest the formulation being modeled? provenance
  2. Does the supplied wording reproduce that passage, translate it, or summarize it? definition

Formulation lineage

Track changes without confusing publication history with fictional history.

Version and attribution

Separate publication evidence, authorship accounts, fictional attribution, and subsequent modifications.

  1. Which evidence establishes this formulation's publication date and distinguishes it from an in-universe attribution? provenance
  2. Which wording changes alter an obligation or exception enough to require a distinct variant record? boundary
Ordered obligations Represent the three duties and their asymmetric conflict relations.

The identity of the Three Laws depends on conditional obligations and precedence, not merely three desirable outcomes.

Human protection

Distinguish direct injury from harm that intervention might prevent.

Action and inaction

Record the proposed interpretation of the protection obligation, including knowledge and intervention assumptions.

  1. What counts as causing injury, and what counts as permitting harm through inaction in this formulation? definition
  2. What must the robot know and be able to do before a failure to intervene is assessed as inconsistent with the First Law? boundary

Subordinate duties

Represent obedience and self-preservation with their higher-priority constraints.

Priority-dependent permission

Trace the obligation that blocks a command or overrides self-protection, while leaving unresolved choices explicit.

  1. Which predicted human harm would make obedience to this command inconsistent with the First Law? action
  2. When self-preservation conflicts with obedience or human protection, which higher-priority obligation applies under the stated facts? action
Interpretive boundaries Expose the meanings required to apply the laws to a particular situation.

The hierarchy alone cannot settle who qualifies as human, which consequences constitute harm, or what speech counts as an order.

Humans and harm

Specify protected subjects and the consequences considered in assessing protection.

Protection semantics

Record human-recognition criteria, harm categories, and prediction horizons as explicit interpretive assumptions.

  1. What evidence or criterion identifies a being as human for this interpretation, and how are uncertain cases represented? definition
  2. Which kinds of harm, affected people, and time horizons enter the assessment? boundary

Orders and existence

Clarify the objects of obedience and self-preservation.

Command and survival meaning

Make assumptions about valid instructions and the robot's continued existence inspectable.

  1. What distinguishes a human order from a suggestion, quotation, deceptive message, or conflicting instruction? definition
  2. Does protecting existence include avoiding shutdown, memory loss, replacement, or physical destruction in the selected interpretation? boundary
Scenario reasoning Assess proposed actions through explicit scenarios, assumptions, and unresolved conflicts.

An agent must distinguish what follows from the laws from what follows only after adding decision rules.

Conflict cases

Identify competing obligations and gaps left by the priority structure.

Underdetermined choices

Record cases involving multiple endangered humans, incompatible commands, or no established harm-free option.

  1. Do the laws resolve this conflict by priority, or does it involve competing demands at the same priority? boundary
  2. What additional tie-breaker or interpretation would be needed to choose an action, and where would it come from? action

Assessment evidence

Keep predicted consequences, narrative evidence, and analyst conclusions distinct.

Conditional consistency

Attach each scenario judgment to the selected formulation, available knowledge, feasible actions, and consequence assumptions.

  1. Which facts about perception, feasible intervention, and predicted outcomes support this consistency judgment? provenance
  2. Does the judgment change when the robot's knowledge or an uncertain consequence changes? action
Extensions and application Control how the three-law system is linked to variants and used outside its fictional setting.

An agent needs to detect changed obligations and distinguish conceptual use from demonstrated operational behavior.

Variant comparison

Compare adaptations and extensions at the level of protected subjects, obligations, and precedence.

Extension deltas

Record what an extension adds or changes without silently replacing the registered three-law referent.

  1. Does this variant add humanity as a protected collective, omit inaction, or change the ordering of duties? boundary
  2. Which source establishes the modification and its relationship to the Three Laws? provenance

Operational use

Assess claims that the laws guide or govern an actual system.

Application claim limits

Distinguish analogy, proposed formalization, implementation claims, and tested behavior.

  1. Is the proposed use literary interpretation, ethical discussion, a formal model, or a claim about implemented robot behavior? definition
  2. What definitions, conflict-resolution rules, and behavioral evidence are needed before accepting the implementation claim? action
Evidence and external alignment What the world already says about this thing, gathered so the model can be checked against it.

A model that cannot be lined up against existing standards, identifiers and practice cannot be adopted by anyone who already uses them.

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.

  • This entry describes a fictional rule system, not a discipline, scientific law or enacted regulation.
  • The familiar explicit formulation appeared in Asimov's 1942 story Runaround; exact wording and later variations should be checked against the relevant published text.
  • This description is based on recall, without source consultation.
  1. Which of these check these first hold for the sense of Three Laws of Robotics this model covers, and on what evidence? provenance

Real-world use

Recalled without web access and unsourced; every item is a lead to verify.

  • Narrative rules used to generate and explore ethical dilemmas in Asimov's robot fiction.
  • Teaching examples for discussing conflicts among safety, obedience and self-preservation.
  • Cultural reference points in debates about robot ethics and AI safety.
  1. Which of these real-world use hold for the sense of Three Laws of Robotics 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.

  • Terms such as harm and human require interpretations that the laws do not operationally specify.
  • Preventing harm through inaction requires predictions about consequences that may be uncertain or unavailable.
  • Conflicting human interests or orders can create dilemmas that the priority hierarchy alone does not resolve.
  • Treating fictional constraints as an implementable safety specification can create unjustified confidence in real systems.
  1. Which of these failure modes and hazards hold for the sense of Three Laws of Robotics 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.

  • Zeroth Law of Robotics - A later extension in Asimov's fiction that places protection of humanity above protection of individual humans; it is not one of the original three laws.
  • Robot ethics - A field of ethical inquiry concerning robots and their development and use; the Three Laws are one fictional formulation discussed within it.
  • AI alignment - Concerns making AI systems behave consistently with intended goals or values; the Three Laws supply a fictional hierarchy rather than a technical alignment method.
  • Robot safety standards - Specify requirements for real robotic systems and their use; Asimov's laws are literary constructs without standards-body authority.
  1. Which of these neighbouring kinds and how to tell them apart hold for the sense of Three Laws of Robotics this model covers, and on what evidence? provenance

What the second pass must settle

  • Which primary edition and passage should anchor the baseline wording, and which translations or revisions produce material differences?
  • What primary evidence clarifies the respective contributions of Asimov and Campbell to the laws' formulation?
  • Which interpretations of human identity, psychological harm, and responsibility for inaction are supported by particular stories rather than generalized across them?
  • Do specific texts supply rules for competing human harms or conflicting human commands, and how narrowly should those rules be scoped?
  • Which proposed formalizations preserve the original priority relations, and which introduce additional assumptions that change the system?