phenomenon · claim-type · cross-domain-transfer · mechanism · claim-relation · replication-status · limitation — grounded in the full text| phenomenon | a field generating ethical declarations faster than it can reconcile them, with 'similar terms often used to mean different things' and a risk the authors name as a 'market for principles'— “Unfortunately, the sheer volume of proposed principles threatens to become overwhelming and confusing, posing two potential problems.1 Either the various sets of ethical principles for AI are similar, leading to unnecessary repetition and redundancy, or, if they differ significantly, confusion and ambiguity will result instead. The worst… | |
| claim-type | method— “Taken together, these five principles capture every one of the 47 principles contained in the six high-profile, expert-driven documents we analysed. Moreover, each principle is included in almost every statement of principles we analyzed” | |
| method — evidence | — “For this comparative analysis, we identified six high-profile initiatives established in the interest of socially beneficial AI: 1. The Asilomar AI Principles, developed under the auspices of the Future of Life Institute” | |
| object of study — evidence | — “Each set of principles meets three basic criteria: they are recent, published within the last three years; directly relevant to AI and its impact on society as a whole (thus excluding documents specific to a particular domain, industry, or sector); and highly reputable, published by authoritative, multi-stakeholder organizations with at least… | |
| research question — evidence | — “Unfortunately, the sheer volume of proposed principles threatens to become overwhelming and confusing, posing two potential problems.1 Either the various sets of ethical principles for AI are similar, leading to unnecessary repetition and redundancy, or, if they differ significantly, confusion and ambiguity will result instead. The worst… | |
| contribution — evidence | — “Taken together, these five principles capture every one of the 47 principles contained in the six high-profile, expert-driven documents we analysed. Moreover, each principle is included in almost every statement of principles we analyzed” | |
| mechanism | the convergence is explained by a structural resemblance between the fields, and the explanation is offered rather than tested: `Of all areas of applied ethics, bioethics is the one that most closely resembles digital ethics in dealing ecologically with new forms of agents, patients, and environments`— “This convergence can most clearly be shown by comparing the sets of principles with the four core principles commonly used in bioethics: beneficence, non-maleficence, autonomy, and justice (Beauchamp & Childress, 2012). The comparison should not be surprising. Of all areas of applied ethics, bioethics is the one that most closely… | mechanisms[0] |
| mechanism | and the fifth principle is argued for FUNCTIONALLY - explicability is needed because the other four cannot be checked without it: you cannot tell whether AI is beneficent without understanding what it does, nor hold anyone to account without knowing why an outcome arose— “The addition of the principle of ‘explicability,’ incorporating both the epistemological sense of ‘intelligibility’ (as an answer to the question ‘how does it work?’) and in the ethical sense of ‘accountability’ (as an answer to the question ‘who is responsible for the way it works?’), is the crucial missing piece of… | mechanisms[1] |
| extends | Beauchamp & Childress's four principles of bioethics, carried into AI ethics with a fifth added and the imperfection of the translation stated Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[0] |
| supports | the 1955 Dartmouth proposal's counterfactual formulation of the AI problem, kept as the working definition sixty-four years later Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[1] |
| contradicts | the claim that ethics is 'Europe's silver bullet' in a 'global AI battle' - answered with two words, 'We disagree', and then with the reason Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[2] |
| sense | `principle proliferation` - THE CONDITION THIS PAPER NAMES AND TREATS. Either the sets are similar, `leading to unnecessary repetition and redundancy`, or they differ, and `confusion and ambiguity will result instead`. The worst case is named too: a `market for principles` in which stakeholders `SHOP` for the most appealing Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[0] |
| sense | THE DIAGNOSIS, WHICH IS THIS PROJECT'S OWN THESIS IN ONE CLAUSE: `the underlying meaning of each of the principles is contested, with similar terms often used to mean different things`. A shared word list is not a shared vocabulary Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[1] |
| sense | explicability - FIVE NAMES FOR ONE PRINCIPLE, LISTED WITH THEIR SOURCES: `TRANSPARENCY` in Asilomar and EGE, `transparency` AND `ACCOUNTABILITY` in IEEE, `INTELLIGIBILITY` in AIUK, `understandable and interpretable` in the Partnership. The new term is coined to cover a thing four documents already had Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[2] |
| sense | explicability, as defined - TWO SENSES DELIBERATELY BOUND TOGETHER: the EPISTEMOLOGICAL sense of intelligibility (`how does it work?`) and the ETHICAL sense of accountability (`who is responsible for the way it works?`). One term carrying two questions on purpose Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[3] |
| sense | AI - `A Research Area in Search of a Definition` in the paper's own heading. It keeps the 1955 Dartmouth formulation, reads it as A COUNTERFACTUAL (`WERE a human to behave in that way, that behaviour WOULD be called intelligent`), and then substitutes its own working phrase - `a reservoir of smart agency on tap` - justified by fitness for the corpus it is about to analyse Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[4] |
| sense | beneficence and non-maleficence - `THOUGH DO ONLY GOOD AND DO NO HARM MAY SEEM LOGICALLY EQUIVALENT, THEY ARE NOT, AND REPRESENT DISTINCT PRINCIPLES`. A near-synonymy refused explicitly, which is exactly the move most papers leave implicit Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[5] |
| sense | meta-autonomy - a coined term for `a DECIDE-TO-DELEGATE model`: humans keep the power to decide WHICH decisions to take, and any delegation must remain `overridable in principle (i.e., deciding to decide again)` Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[6] |
| control | the six documents were selected by three declared criteria - recent, relevant to AI and society as a whole rather than to one sector, and published by authoritative multi-stakeholder bodies of at least national scope — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. — rules out ad-hoc selection as the explanation of the convergence; the criteria are stated before the result and exclude domain-specific documents Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[0] |
| control | whether the convergence is an artefact of where the documents come from — failed Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. — rules out nothing, and the authors say so: all six came from global-scope initiatives or 'from within western liberal democracies', and the framework would benefit from perspectives 'presently un- or under-represented in our sample' Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[1] |
| control | whether the framework's later adoption is independent evidence for it — failed Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. — rules out nothing - AI4People and the EC High-Level Expert Group are counted as uptake, and the disclosure states the authors chaired, rapporteured and sat on them. THE DISCLOSURE IS COMPLETE AND THE CIRCULARITY IS STILL THERE Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[2] |
| control | whether 'beneficence' and 'non-maleficence' are one principle wearing two names — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. — rules out the collapse, explicitly: they 'may seem logically equivalent, they are not, and represent distinct principles', and the six documents treat them separately Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[3] |
| weight | Beauchamp & Childress's four principles of bioethics, carried into AI ethics with a fifth added and the imperfection of the translation stated — 0.9 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[0] |
| weight | the 1955 Dartmouth proposal's counterfactual formulation of the AI problem, kept as the working definition sixty-four years later — 0.7 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[1] |
| weight | the claim that ethics is 'Europe's silver bullet' in a 'global AI battle' - answered with two words, 'We disagree', and then with the reason — 0.8 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[2] |
| edge | beneficence —distinct-from→ non-maleficence Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concept_edges[0] |
| audit | digital ethics —distinct-from→ bioethics — Audit 2026-09-28: DOUBTFUL - read against its span and kept, but the span compares AI principles with bioethics principles; distinct-from is loose. Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concept_edges[1] |
| outcome | control — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[0] |
| outcome | control — failed Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[1] |
| outcome | control — failed Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[2] |
| outcome | control — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[3] |
| limitation | THE SAMPLE IS SIX DOCUMENTS AND THE PAPER SAYS WHERE THEY COME FROM: each `emerged either from initiatives with global scope, or from within western liberal democracies`, and the framework `would undoubtedly benefit from the perspectives of regions and cultures presently un- or under-represented` Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | limitations[0] |
| limitation | the coverage claim and the distribution claim are DIFFERENT CLAIMS AND ONLY THE FIRST IS ABSOLUTE: the five principles `capture every one of the 47`, but each principle appears in `almost every statement` - and the paper's own Table 1 shows IEEE missing autonomy and justice, and the Partnership missing autonomy Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | limitations[1] |
| limitation | THE ADOPTION EVIDENCE INCLUDES BODIES THE AUTHORS SIT ON. AI4People and the EC High-Level Expert Group are counted as uptake, and the disclosure states that FLORIDI CHAIRED AI4PEOPLE, COWLS WAS ITS RAPPORTEUR, AND FLORIDI IS A MEMBER OF THE HLEG. Declared in full, and it is still the authors' framework being adopted by the authors' committees Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | limitations[2] |
| limitation | the funding sits on the subject: the work was supported by an EPSRC hub AND BY FACEBOOK AND GOOGLE - two of the companies whose AI the framework is written to govern. Declared Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | limitations[3] |
| limitation | the paper leaves its central ambiguity open and says so twice: `it is not entirely clear whether it is the people developing AI, or the technology itself, which should be encouraged not to do harm - in other words, whether it is Frankenstein or his monster against whose maleficence we should be guarding` Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | limitations[4] |
| limitation | and `justice` is left carrying more than one meaning inside the framework that is supposed to fix it: `Elsewhere ‘justice’ has still other meanings (especially in the sense of _fairness)`. The unification does not reach inside the principles Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | limitations[5] |
finding · quantity · variable · definition · key-term — grounded in the full text| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[0] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[1] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[2] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[3] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[4] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[5] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[6] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[7] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[8] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[9] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[10] |
| quantity | principles across the six documents, claimed to be captured entirely by the five Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.quantities[0] |
| quantity | high-profile initiatives analysed - Asilomar, Montreal, IEEE, EGE, AIUK, the Partnership on AI Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.quantities[1] |
| quantity | core principles in the resulting framework - four borrowed, one coined Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.quantities[2] |
| definition | explicability Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.definitions[0] |
| definition | artificial intelligence Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.definitions[1] |
| definition | meta-autonomy Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.definitions[2] |
| concept | principle proliferation — not defined in this paper's own words; the condition of a field producing more ethical declarations than it can reconcile Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[0] |
| concept | explicability — not defined in this paper's own words; the fifth principle, coined here to cover intelligibility and accountability together Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[1] |
| concept | intelligibility — not defined in this paper's own words; the epistemological half of explicability - an answer to 'how does it work?' Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[2] |
| concept | accountability — not defined in this paper's own words; the ethical half of explicability - an answer to 'who is responsible for the way it works?' Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[3] |
| concept | beneficence — not defined in this paper's own words; 'do only good' - one of the four principles imported from bioethics Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[4] |
| concept | non-maleficence — not defined in this paper's own words; 'do no harm' - argued to be NOT logically equivalent to beneficence Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[5] |
| concept | bioethics — not defined in this paper's own words; the field the four core principles are imported from, chosen because it 'most closely resembles digital ethics' Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[6] |
| concept | digital ethics — not defined in this paper's own words; the importing field - and the import is stated to be imperfect Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[7] |
| concept | artificial intelligence — not defined in this paper's own words; kept at the 1955 Dartmouth formulation, read as a counterfactual about BEHAVIOUR and explicitly not about thinking Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[8] |
| concept | thinking — not defined in this paper's own words; the mental predicate the Dartmouth counterfactual explicitly does NOT license - 'It does not mean that the machine _is_ intelligent, or even _thinking' Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[9] |
| concept | meta-autonomy — not defined in this paper's own words; a decide-to-delegate model in which delegation stays overridable in principle Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[10] |
When we have changed our mind. Some entries in the table begin with a word in capitals and the reason for it. EVIDENCE CORRECTED — the finding still stands, but we had quoted the wrong sentence from the paper, and the quotation has been replaced with the right one. RETRACTED — the entry itself was wrong; it stays on the page, marked, with the reason, rather than being deleted. WITHDRAWN AND REPLACED — the entry was wrong, and a corrected version of it appears in the entries below. Each mark carries its own reason, in brackets, because the reasons differ from one entry to the next. An entry we have WITHDRAWN is removed from this page and kept in the record, with its reason. A line reading Audit and a date is a later re-reading of the entry against its whole sentence: DOUBTFUL means it was kept, and the reason says what is loose about it. Quotation marks inside the tagger's sentence. Where a tag is the tagger's own sentence, a phrase in quotation marks inside it is borrowed from the paper — the tagger's wording around it is not. The paper's own sentence is the evidence quote for that row; on this page it is withheld with the article, and the article is at its publisher. Nothing inside quotation marks anywhere on this page is ours: if it is quoted, it is the paper's. Whose words a concept carries. A concept entry records a term the paper itself uses, so the term, the sense given for it, and its evidence are the paper's own words, quoted where the paper fixes them. This is the opposite of a sense entry in the tag table, which is the tagger's sentence about how the paper uses a term. Where a concept entry could not be grounded in the paper's words, the entry says so on its own row.
Tag layer © the Meta-Tagging Project, CC BY 4.0. The paper text is © its authors/publisher
and is linked, not reproduced. Excerpts shown are short quotations for identification and verification.
· back to index
· concepts · authors
· עברית
© 2026 Shir Sivroni · ORCID 0009-0000-2597-2824