Artificial General Ethics: Toward Auditable Moral Reasoning Across Domains

Image
Artificial General Ethics Image
Loading voting controls…
  • Large-scale empirical moral preference research (Emerging Research): Cross-cultural experiments reveal both shared patterns and substantial variation in how people judge machine dilemmas, warning against a single hidden value model.

  • Machine moral judgment benchmarks (Experimental): Recent systems can classify or generate judgments in bounded tasks, but consistency with survey labels is not the same as justified moral reasoning.

  • Language-model moral reasoning evaluation (Experimental): Benchmark studies probe whether models can explain, generalize and remain coherent under changed premises instead of merely matching familiar answers.

  • Principle-guided model training (Experimental): Constitutional and rule-guided approaches show that explicit principles can shape model behavior, while leaving open who selects, interprets and contests those principles.

  • The integrated field is classified as Hypothetical. Its decisive unknowns include representing moral pluralism—Systems need to preserve legitimate disagreement and jurisdictional context without collapsing every conflict into one numerical objective; reason-giving under uncertainty—A model must expose facts, affected parties, assumptions, principles, counterarguments and confidence rather than only output a verdict; cross-domain transfer without moral drift—Ethical competence in one benchmark must not be assumed to transfer to medicine, law, war, education or intimate life.

Table of contents

Current section:

Introduction to Artificial General Ethics

Artificial general ethics is a proposed interdisciplinary science for building and evaluating AI systems that can identify moral stakes, reason across competing values and remain accountable across unfamiliar domains.

Its purpose is not to create a machine that declares a universal morality, but to develop transparent, contestable and culturally plural methods for ethical reasoning that remain subordinate to human rights and legitimate institutions. Its present evidence level is Hypothetical: the field is neither described as a completed discipline nor reduced to a fantasy because its final instruments do not yet exist.

What is Artificial General Ethics?

Artificial general ethics is a proposed interdisciplinary science for building and evaluating AI systems that can identify moral stakes, reason across competing values and remain accountable across unfamiliar domains.

Future Sciences assumes that humanity will continue inventing disciplines for questions current fields cannot yet answer; the task of this article is to make that possibility researchable rather than merely inspirational. The practical bridge begins with large-scale empirical moral preference research, machine moral judgment benchmarks, and language-model moral reasoning evaluation. Those foundations already provide measurements, models or prototypes from which a distinct research community could grow.

The destination is intentionally ambitious: a public science of machine-assisted moral reasoning whose systems can explain their assumptions, preserve pluralism, submit to appeal and strengthen rather than replace accountable human institutions. No calendar can responsibly promise this destination. Progress can still be recognized whenever Artificial General Ethics converts one unknown—beginning with representing moral pluralism—into a reproducible capability.

Artificial General Ethics should be understood as a proposed scientific integration, not merely a new label for one existing specialty. Its identity comes from a particular objective: not to create a machine that declares a universal morality, but to develop transparent, contestable and culturally plural methods for ethical reasoning that remain subordinate to human rights and legitimate institutions.

A future community must be able to reproduce ethical issue detection, audit moral authority capture and distinguish an engineering setback from a falsified scientific premise. Current disciplines can supply components, but a mature Artificial General Ethics would connect them into a reproducible program directed toward a public science of machine-assisted moral reasoning whose systems can explain their assumptions, preserve pluralism, submit to appeal and strengthen rather than replace accountable human institutions.

This distinction matters for search readers and researchers alike. The article separates what can be done now, what exists only in bounded experiments, what remains hypothetical and what belongs to the deepest horizon. The page therefore protects the ambition of Artificial General Ethics without presenting tomorrow's achievement as today's evidence.

Artificial General Ethics is not a claim that every enabling technology is mature. It is a bounded research identity: a defined problem, a set of inherited methods, explicit exclusions and measurable conditions under which the field could advance or fail.

Why Artificial General Ethics matters for humanity

Future sciences become necessary when established specialties can describe pieces of a problem but no single discipline can organize the whole journey. Artificial general ethics is a proposed interdisciplinary science for building and evaluating AI systems that can identify moral stakes, reason across competing values and remain accountable across unfamiliar domains.

A credible program could advance ethical issue detection and deliberation support while building the measurement standards required for policy stress testing. The aim is cumulative capability, not novelty for its own sake.

Civilizational value and scientific restraint must grow together. Because moral authority capture could undermine the very purpose of the field, progress must be judged by safety, distribution of benefits and the quality of human oversight as well as technical performance.

Scientific foundations and historical path

Parent disciplines and their contributions

ComponentEvidence levelWhat is supported todayWhat remains to be achieved
Large-scale empirical moral preference researchEmerging ResearchCross-cultural experiments reveal both shared patterns and substantial variation in how people judge machine dilemmas, warning against a single hidden value model.Representing moral pluralism
Machine moral judgment benchmarksExperimentalRecent systems can classify or generate judgments in bounded tasks, but consistency with survey labels is not the same as justified moral reasoning.Representing moral pluralism
Language-model moral reasoning evaluationExperimentalBenchmark studies probe whether models can explain, generalize and remain coherent under changed premises instead of merely matching familiar answers.Representing moral pluralism
Principle-guided model trainingExperimentalConstitutional and rule-guided approaches show that explicit principles can shape model behavior, while leaving open who selects, interprets and contests those principles.Representing moral pluralism
Integrated Artificial General EthicsHypotheticalThe field has a coherent objective and identifiable enabling sciences.A validated integration that advances toward a public science of machine-assisted moral reasoning whose systems can explain their assumptions, preserve pluralism, submit to appeal and strengthen rather than replace accountable human institutions.

Overall classification: The proposed discipline is classified as Hypothetical: scientifically formulable and connected to present foundations, but not yet unified as the proposed discipline. Its component foundations span Emerging Research, Experimental. The field-level rating must not downgrade established tools or upgrade representing moral pluralism before it is demonstrated.

Historical milestones

The field does not begin with its new name. It inherits a sequence of discoveries and institutions that progressively made its central questions measurable.

  1. 2001: Convention on Cybercrime (Budapest Convention) . Council of Europe (2001; current treaty framework). Primary or institutional source .
  2. 2018: The Moral Machine experiment . Nature (2018). Primary or institutional source .
  3. 2021: Recommendation on the Ethics of Artificial Intelligence . UNESCO (2021). Primary or institutional source .
  4. 2022: Constitutional AI: Harmlessness from AI Feedback . Anthropic research (2022). Primary or institutional source .

These milestones establish a path into Artificial General Ethics; none alone demonstrates that the integrated future science already exists.

Why this field is emerging now

Artificial General Ethics is becoming researchable now because the cited component sciences can increasingly measure, model or prototype parts of its central problem. The convergence is scientifically meaningful only where those components can be integrated without erasing their different evidence levels and limitations.

Current scientific advances that point toward this field

Landmark foundations

The most important signals are not promises of a completed discipline. They are reproducible results in neighboring fields that expose mechanisms, instruments and limits the future science can inherit.

The first bridge into Artificial General Ethics is built from evidence that already has methods, data and institutions. The most defensible starting points for Artificial General Ethics are the following lines of work, each with a different evidence level and a different role in the proposed discipline.

Recent advances

These institutes connect law, philosophy, computation and public institutions, helping define not only what a system can do but who may challenge it and under which authority.

Legal-technology platforms show how computational tools enter professional practice, while also making opacity, vendor dependence and procedural accountability measurable concerns.

What these advances do not yet prove

These results do not by themselves establish the integrated Artificial General Ethics discipline. They support bounded mechanisms, instruments or prototypes. Claims of transfer, superiority, safety or social benefit require direct comparison with mature alternatives and independent replication at the scale of the intended application.

Research ecosystem: universities, laboratories, industry, and institutions

Universities, laboratories, and research centers

  • Named institutions and their specific programs are documented in the cited source record and require human verification.

Industry and applied innovation

  • Applied actors must be assessed through independently verifiable programs rather than marketing claims.

Standards, regulators, and multilateral bodies

Frontier status: evidence and maturity

What is already established

No integrated version of Artificial General Ethics is established. Its strongest present foundations are separately recognized methods and observations, especially large-scale empirical moral preference research. The evidence belongs to these components at their demonstrated scale; it does not automatically validate the proposed synthesis.

What is emerging

large-scale empirical moral preference research—Cross-cultural experiments reveal both shared patterns and substantial variation in how people judge machine dilemmas, warning against a single hidden value model.; machine moral judgment benchmarks—Recent systems can classify or generate judgments in bounded tasks, but consistency with survey labels is not the same as justified moral reasoning.; language-model moral reasoning evaluation—Benchmark studies probe whether models can explain, generalize and remain coherent under changed premises instead of merely matching familiar answers. These lines of work create an experimental bridge, but transfer across laboratories, populations and operating conditions remains a central test.

What remains hypothetical or speculative

The integrated field is classified as Hypothetical. Its decisive unknowns include representing moral pluralism—Systems need to preserve legitimate disagreement and jurisdictional context without collapsing every conflict into one numerical objective.; reason-giving under uncertainty—A model must expose facts, affected parties, assumptions, principles, counterarguments and confidence rather than only output a verdict.; cross-domain transfer without moral drift—Ethical competence in one benchmark must not be assumed to transfer to medicine, law, war, education or intimate life. The long-term destination—a public science of machine-assisted moral reasoning whose systems can explain their assumptions, preserve pluralism, submit to appeal and strengthen rather than replace accountable human institutions—is a research horizon, not a forecast or current capability.

Evidence map

ComponentCurrent evidenceWhat remains unresolved
Large-scale empirical moral preference researchCross-cultural experiments reveal both shared patterns and substantial variation in how people judge machine dilemmas, warning against a single hidden value model.Independent transfer, causal attribution and field-level validation remain necessary before this component can support the complete Artificial General Ethics capability.
Machine moral judgment benchmarksRecent systems can classify or generate judgments in bounded tasks, but consistency with survey labels is not the same as justified moral reasoning.Independent transfer, causal attribution and field-level validation remain necessary before this component can support the complete Artificial General Ethics capability.
Language-model moral reasoning evaluationBenchmark studies probe whether models can explain, generalize and remain coherent under changed premises instead of merely matching familiar answers.Independent transfer, causal attribution and field-level validation remain necessary before this component can support the complete Artificial General Ethics capability.
Principle-guided model trainingConstitutional and rule-guided approaches show that explicit principles can shape model behavior, while leaving open who selects, interprets and contests those principles.Independent transfer, causal attribution and field-level validation remain necessary before this component can support the complete Artificial General Ethics capability.

Fundamental principles of Artificial General Ethics

The discipline should be built around causal mechanisms, explicit uncertainty, open comparison and failure criteria. The following breakthroughs are not decorative forecasts; they are the scientific conditions required for the field to become distinct and cumulative.

  • Representing moral pluralism — Systems need to preserve legitimate disagreement and jurisdictional context without collapsing every conflict into one numerical objective. Progress should be measured by a preregistered benchmark, independent replication and a clear account of what result would invalidate the proposed approach.
  • Reason-giving under uncertainty — A model must expose facts, affected parties, assumptions, principles, counterarguments and confidence rather than only output a verdict. Progress should be measured by a preregistered benchmark, independent replication and a clear account of what result would invalidate the proposed approach.
  • Cross-domain transfer without moral drift — Ethical competence in one benchmark must not be assumed to transfer to medicine, law, war, education or intimate life. Progress should be measured by a preregistered benchmark, independent replication and a clear account of what result would invalidate the proposed approach.
  • Institutional accountability — No ethical model is sufficient without appeal, audit, human authority, monitoring and the ability to revise governing principles. The breakthrough is scientific only when it changes prediction, measurement or control in a way that competing methods cannot match.

Methods, tools, data, and validation

Methods and instruments

Comparable protocols are the mechanism by which Artificial General Ethics can separate robust effects from laboratory-specific demonstrations. The methods below translate the mission into an experimental architecture.

Doctrinal and computational analysis

Link machine-readable rules and empirical outcomes to constitutional principles, institutional competence and existing sources of law. Evaluation must include technical performance, transfer across contexts and the social or biological outcome the system is meant to improve.

Procedural benchmark design

Measure notice, explanation, contestability, equality of arms, evidentiary reliability and remedy—not only prediction accuracy. A shared protocol would let independent laboratories compare results without requiring identical hardware, populations or institutional assumptions.

Regulatory sandboxes with sunset clauses

Allow bounded experimentation while requiring logs, external review, rollback and automatic expiration unless benefits are demonstrated. The method should expose uncertainty and preserve negative results, because the field cannot mature if only successful prototypes enter its record.

Comparative legal stress testing

Examine how a proposal behaves across jurisdictions, cultures, emergencies and asymmetric power relationships. The method should expose uncertainty and preserve negative results, because the field cannot mature if only successful prototypes enter its record.

Data, models, and benchmarks

Data architecture for Artificial General Ethics must preserve provenance, uncertainty, population or environmental context, negative results and the distinction between measured variables and model-generated inference. Benchmarks should compare the proposed method with the strongest established alternative on the same task.

Validation, replication, and falsification

Validation requires preregistered hypotheses, independent replication, out-of-distribution testing and an explicit result that would falsify the central mechanism. A component-level gain is not a field-level advantage unless it changes the intended scientific or public outcome after cost, error, safety and downstream processing are included.

Breakthroughs still required

Representing moral pluralism

Systems need to preserve legitimate disagreement and jurisdictional context without collapsing every conflict into one numerical objective. Progress should be measured by a preregistered benchmark, independent replication and a clear account of what result would invalidate the proposed approach.

Measurable success criterion: Success would require a preregistered, independently reproduced test of representing moral pluralism that demonstrates this condition under realistic settings for Artificial General Ethics: Systems need to preserve legitimate disagreement and jurisdictional context without collapsing every conflict into one numerical objective. Failure criterion: The pathway should be revised or rejected if the effect disappears under stronger controls, fails to transfer, or is matched by a safer conventional method.

Reason-giving under uncertainty

A model must expose facts, affected parties, assumptions, principles, counterarguments and confidence rather than only output a verdict. Progress should be measured by a preregistered benchmark, independent replication and a clear account of what result would invalidate the proposed approach.

Measurable success criterion: Success would require a preregistered, independently reproduced test of reason-giving under uncertainty that demonstrates this condition under realistic settings for Artificial General Ethics: A model must expose facts, affected parties, assumptions, principles, counterarguments and confidence rather than only output a verdict. Failure criterion: The pathway should be revised or rejected if the effect disappears under stronger controls, fails to transfer, or is matched by a safer conventional method.

Cross-domain transfer without moral drift

Ethical competence in one benchmark must not be assumed to transfer to medicine, law, war, education or intimate life. Progress should be measured by a preregistered benchmark, independent replication and a clear account of what result would invalidate the proposed approach.

Measurable success criterion: Success would require a preregistered, independently reproduced test of cross-domain transfer without moral drift that demonstrates this condition under realistic settings for Artificial General Ethics: Ethical competence in one benchmark must not be assumed to transfer to medicine, law, war, education or intimate life. Failure criterion: The pathway should be revised or rejected if the effect disappears under stronger controls, fails to transfer, or is matched by a safer conventional method.

Institutional accountability

No ethical model is sufficient without appeal, audit, human authority, monitoring and the ability to revise governing principles. The breakthrough is scientific only when it changes prediction, measurement or control in a way that competing methods cannot match.

Measurable success criterion: Success would require a preregistered, independently reproduced test of institutional accountability that demonstrates this condition under realistic settings for Artificial General Ethics: No ethical model is sufficient without appeal, audit, human authority, monitoring and the ability to revise governing principles. Failure criterion: The pathway should be revised or rejected if the effect disappears under stronger controls, fails to transfer, or is matched by a safer conventional method.

Research roadmap

Stage 1 — definitions, baselines, and open data

Define the field’s objects and exclusions, preserve the strongest existing evidence, publish baseline datasets and establish where current methods fail.

Stage 2 — measurement and causal models

Develop measurements for Representing moral pluralism and compare causal explanations prospectively rather than fitting a preferred story after the result.

Stage 3 — bounded experimental systems

Test Reason-giving under uncertainty in reversible prototypes with explicit stop conditions, strong comparators and monitoring of unintended effects.

Stage 4 — independent validation and responsible scale

Require multi-site replication, standards, security, governance and evidence that Cross-domain transfer without moral drift survives heterogeneous real-world conditions.

Stage 5 — long-term scientific capability

Integrate only validated components into a mature Artificial General Ethics capability, while preserving human authority, reversibility and the ability to abandon failed mechanisms.

Potential applications

Current and adjacent applications

Applications should be staged by evidence and dependency. Near-term work extends existing methods; long-term possibilities require integration; transformative scenarios depend on discoveries that may take generations.

Near- and mid-term applications

If the research program succeeds, Artificial General Ethics could contribute to ethical issue detection, deliberation support, policy stress testing and adjacent missions. The list is an agenda for bounded trials and long-term validation rather than a catalogue of existing services.

Long-term possibilities

Long-term applications depend on the breakthroughs and validation stages defined above.

Transformative scenarios

Transformative uses of Artificial General Ethics remain conditional scenarios and should never be represented as present services or guaranteed outcomes.

Ethical, legal, safety, and human challenges

Future law must preserve due process, human dignity and meaningful remedy even when evidence, actors or environments are technologically unfamiliar. Efficiency is not a substitute for legitimacy, and prediction is not judgment.

Moral authority capture

Governments or vendors may present one system's outputs as neutral ethics while embedding their own interests. Before Artificial General Ethics scales, independent evaluators should publish known failure modes related to moral authority capture.

Value homogenization

Training on dominant languages and institutions can erase minority traditions and legitimate disagreement. Design should reduce the technical pathway to moral authority capture instead of depending only on promises made after deployment.

Responsibility laundering

Organizations can blame an ethical model for decisions that remain human and institutional choices. People affected by Artificial General Ethics need notice, participation, a way to contest outcomes and an effective remedy.

Persuasive moral manipulation

Systems able to model values may be optimized to pressure users rather than support autonomous deliberation. Lifecycle monitoring is essential because consequences of ethical issue detection may appear after the bounded trial has ended.

Ethical architecture must evolve alongside large-scale empirical moral preference research; it cannot be postponed until the technology reaches ethical issue detection. For a capability as consequential as Artificial General Ethics, consent, distribution of benefit, reversibility, accountability and long-term monitoring determine which experiments are scientifically acceptable in the first place.

Societal and civilizational outlook

A dependency-based roadmap protects Artificial General Ethics from declaring maturity because one prototype appears on schedule. A later stage should not be declared complete because a product uses the field's name; it should inherit evidence from the stages beneath it.

Define the objects, outcomes and exclusions of Artificial General Ethics. Build datasets and baseline methods from large-scale empirical moral preference research and machine moral judgment benchmarks, documenting where current approaches fail.

Develop instruments that can observe the variables implied by representing moral pluralism. Compare competing mechanisms prospectively and publish null results so that the field does not grow around untested assumptions.

Construct reversible prototypes for ethical issue detection and deliberation support. Trials should begin in controlled settings with explicit stop conditions, independent monitoring and strong conventional comparators.

Create specialist training, replication networks, shared standards and governance able to address moral authority capture and value homogenization. A field at this stage would have results that transfer across laboratories and populations.

Integrate the validated components until humanity can pursue a public science of machine-assisted moral reasoning whose systems can explain their assumptions, preserve pluralism, submit to appeal and strengthen rather than replace accountable human institutions. The final stage has no responsible fixed date: it advances when prerequisite discoveries are demonstrated, not when a forecast expires.

The horizon that gives coherence to Artificial General Ethics is a public science of machine-assisted moral reasoning whose systems can explain their assumptions, preserve pluralism, submit to appeal and strengthen rather than replace accountable human institutions. That destination may sit far beyond current laboratories, but it clarifies why the field is worth defining: present researchers can identify prerequisites, build instruments and prevent future generations from inheriting a powerful capability with no scientific or ethical architecture.

Future Sciences does not require every proposed mechanism inside Artificial General Ethics to survive. It is that humanity can continue expanding the domain of the scientifically knowable. The correct response to a missing method is therefore a better question, a discriminating experiment and a roadmap that can survive the replacement of today's theories.

Scientific maturity arrives when the field's predictions are riskier than its rhetoric and its failures are publicly legible. Until then, Artificial General Ethics remains a disciplined invitation to build the science its goal requires.

The civilizational value of Artificial General Ethics should be judged through distribution of benefits, resilience, reversibility and the quality of institutions able to challenge the technology. A future capability is not progress if its gains depend on hidden externalities, coerced participation or the loss of meaningful human or ecological agency.

Learning path to master Artificial General Ethics

No university degree is yet required to carry the exact name Artificial General Ethics. The responsible path is to become excellent in recognized disciplines, then use the proposed field to define an interdisciplinary research question.

Undergraduate foundations

Students should build mathematical literacy, experimental discipline and domain knowledge before specializing in the future integration.

  • Law
  • Political Science
  • Computer Science
  • Statistics
  • Philosophy

Graduate studies

Students should build mathematical literacy, experimental discipline and domain knowledge before specializing in the future integration.

  • Law
  • Political Science
  • Computer Science
  • Statistics
  • Philosophy

PhD-level research

A doctoral project should contribute one falsifiable bridge rather than claim to complete the entire future science.

  • Learn to formalize contestable legal reasoning in the context of Artificial General Ethics.
  • Learn to design procedural benchmarks in the context of Artificial General Ethics.
  • Learn to evaluate institutional feedback in the context of Artificial General Ethics.
  • Learn to compare governance across jurisdictions in the context of Artificial General Ethics.

Core skills, methods, and tools

The most useful curriculum combines the following areas with scientific writing, open methods, ethics and collaboration across institutions.

  • Jurisprudence
  • Administrative Law
  • Machine Learning
  • Cybersecurity
  • Research Methods
  • Ethics
  • Public Administration

Careers and fields of contribution

Existing roles that can contribute today

Most contributors will initially work under established professional titles rather than as “Artificial General Ethics scientists.” That is normal: a future discipline becomes real when specialists learn to coordinate around shared questions, datasets and standards.

Universities can contribute through interdisciplinary laboratories and doctoral programs; industry through transparent engineering and benchmark participation; governments through public-interest research, standards and oversight; and civil society through rights, community knowledge and independent scrutiny. The field should reward people who publish limitations and negative results, not only spectacular demonstrations.

  • Computational Legal Researcher — contributes methods, evidence or governance to one part of the emerging discipline.
  • Ai Governance Counsel — contributes methods, evidence or governance to one part of the emerging discipline.
  • Digital-Evidence Specialist — contributes methods, evidence or governance to one part of the emerging discipline.
  • Regulatory Technologist — contributes methods, evidence or governance to one part of the emerging discipline.
  • Public-Interest Algorithm Auditor — contributes methods, evidence or governance to one part of the emerging discipline.
  • Future-Law Scholar — contributes methods, evidence or governance to one part of the emerging discipline.

Possible future roles

Possible future roles should be named only after the discipline develops recognized methods, training and accountability. They may include a Artificial General Ethics research scientist, field-specific validation lead, safety and governance specialist, or interdisciplinary program director. These are projected roles, not current standardized occupations.

Open questions for future researchers

Scientific identity emerges from problems whose answers can surprise every side; Artificial General Ethics now needs that kind of agenda. The following questions form an initial agenda for Artificial General Ethics.

  1. Which observation would distinguish Artificial General Ethics from the best existing approach in law, evidence and future governance?
  2. How can large-scale empirical moral preference research and machine moral judgment benchmarks be connected without overstating what either currently proves?
  3. What experiment would falsify the central assumption behind representing moral pluralism?
  4. Which benchmark would show that ethical issue detection has improved a real outcome rather than a proxy?
  5. How can researchers prevent moral authority capture while preserving the capability the field is meant to create?
  6. Which parts of the system must remain reversible, interruptible or under direct human authority?
  7. Who should control the data, instruments and infrastructure needed to develop Artificial General Ethics?
  8. What discovery would justify moving the discipline from Hypothetical to the next evidence level?

Frequently asked questions

What is Artificial General Ethics?

Artificial general ethics is a proposed interdisciplinary science for building and evaluating AI systems that can identify moral stakes, reason across competing values and remain accountable across unfamiliar domains.

Does Artificial General Ethics already exist?

The integrated field is classified as Hypothetical. Its component sciences and technologies exist at different maturity levels, but the complete discipline should not be treated as established unless the evidence section explicitly says so.

What evidence supports it?

Large-scale empirical moral preference research (Emerging Research): Cross-cultural experiments reveal both shared patterns and substantial variation in how people judge machine dilemmas, warning against a single hidden value model.

What breakthrough matters most?

Representing moral pluralism: Systems need to preserve legitimate disagreement and jurisdictional context without collapsing every conflict into one numerical objective. Progress should be measured by a preregistered benchmark, independent replication and a clear account of what result would invalidate the proposed approach.

How can someone study or contribute to it?

Begin with recognized programs in Law, Political Science, Computer Science, Statistics, Philosophy. Then define a falsifiable interdisciplinary question, work with domain specialists and publish both positive and negative results.

Related Future Sciences

These related sciences represent enabling disciplines, shared risks or downstream capabilities. Links are included only where the relationship is scientifically meaningful.

References and further reading

The evidence base below explains why Artificial General Ethics can be formulated scientifically while preserving uncertainty about its mature form.

  1. The Moral Machine experiment. Nature (2018). Primary or institutional source.
  2. Investigating machine moral judgement through the Delphi experiment. Nature Machine Intelligence (2025). Primary or institutional source.
  3. Rethinking Machine Ethics: Can Language Models Perform Moral Reasoning?. Findings of NAACL (2024). Primary or institutional source.
  4. Ethical Reasoning over Moral Alignment. Findings of EMNLP (2023). Primary or institutional source.
  5. Constitutional AI: Harmlessness from AI Feedback. Anthropic research (2022). Primary or institutional source.
  6. Measuring Large Language Models' Alignment with Utilitarian Moral Dilemmas. EMNLP (2024). Primary or institutional source.
  7. Artificial Intelligence Risk Management Framework (AI RMF 1.0). NIST (2023). Primary or institutional source.
  8. Recommendation on the Ethics of Artificial Intelligence. UNESCO (2021). Primary or institutional source.
  9. Regulation (EU) 2024/1689 — Artificial Intelligence Act. European Union (2024). Primary or institutional source.
  10. Post-Quantum Cryptography Standards. NIST (2024–2026). Primary or institutional source.
  11. Institute for Ethics in AI. University of Oxford (ongoing). Primary or institutional source.
  12. Research at the Stanford Institute for Human-Centered Artificial Intelligence. Stanford HAI (ongoing). Primary or institutional source.
  13. Research at MIT Computer Science and Artificial Intelligence Laboratory. MIT CSAIL (ongoing). Primary or institutional source.
  14. Artificial Intelligence Research. Microsoft Research (ongoing). Primary or institutional source.
  15. Convention on Cybercrime (Budapest Convention). Council of Europe (2001; current treaty framework). Primary or institutional source.

Evidence level: Hypothetical. Review status: Specialist scientific review pending.

Editorial disclosure: AI contributed to research organization and prose generation. Publication responsibility, including fact-checking and evidence classification, remains with the Future Sciences editorial team.

Evidence level: Hypothetical. Review status: Human scientific and journalistic review required before publication.

Editorial disclosure: AI tools assisted with corpus comparison, structural normalization and drafting. Human editors and domain specialists remain responsible for verifying every claim, source interpretation, link and field-specific term.

Explore, Discover, Transcend

Artificial General Ethics will not be founded by a title alone. It will emerge when researchers can connect evidence, instruments, criticism and purpose across disciplines while remaining honest about every unknown.

Artificial General Ethics draws meaning from adjacent future sciences. These relationships represent enabling knowledge, shared risks or capabilities that may emerge downstream.

Future Sciences invites the next generation to study the foundations, challenge the assumptions and invent the missing methods. The destination is a public science of machine-assisted moral reasoning whose systems can explain their assumptions, preserve pluralism, submit to appeal and strengthen rather than replace accountable human institutions. The first step is a question precise enough to test today.

Lineage compass

Scientific genealogy

Reviewed direct foundations converging into this Science.

Historical reference

Philosophy

Contribution
Theoretical
Evidence level
Speculative

Current Science

Artificial General Ethics: Toward Auditable Moral Reasoning Across Domains

The Science you are reading

Past / Present / Future

Science trajectory

Follow this Science and its evidence-backed parent lineage from origin to estimated practical use and maturity. The present starts centered; use Focus now to return to the current year.

  • X · TimeEach division uses the selected number of years. The present starts centered; drag horizontally to review each Science from origin to maturity.
  • Y · Development stageOrigin, practical use and peak maturity form one trajectory.
  • Origin rangeThe horizontal bar shows uncertainty; future dates are editorial scenarios.

Use Tab and the arrow keys to focus a Science, Enter to open its evidence, Escape to close details, drag horizontally to review the full trajectory, and Focus now to restore the present.

Science trajectory Interactive genealogy centered on the current year. A complete text equivalent follows the diagram.
Mathematics 2750 BCE
Computer Science 1946 CE
Philosophy 550 BCE
Artificial Intelligence 1956 CE
Artificial General Ethics: Toward Auditable Moral Reasoning Across Domains 2026 CE
Browse all genealogy data and sources
  1. Ancestor generation 1

  2. Ancestor generation 2

    • Computer Science

      Origin
      1936 CE - 1956 CE
      High confidence
      Formal models of computation and early stored-program machines established the basis of modern computer science.
      Evidence level: Established Science
      Editorial publication assisted by AI/MCP.
      Practical Use
      1956 CE - 1990 CE
      High confidence
      Computing became an academic discipline and operational technology across science, government and industry.
      Evidence level: Established Science
      Editorial publication assisted by AI/MCP.
      Peak
      1990 CE - 2026 CE
      High confidence
      Networked computing, large-scale software and machine learning made computer science a pervasive enabling discipline.
      Evidence level: Established Science
      Editorial publication assisted by AI/MCP.
      • Technological contribution to Artificial Intelligence

        Computer Science contributes established concepts and methods to Artificial Intelligence. This reviewed edge records documented disciplinary inheritance without reducing either field to a single origin.

        Evidence level: Established Science

        Editorial publication assisted by AI/MCP.

  3. Ancestor generation 3

  4. Current Science

Comments