standOff, and the article is at its publisher via the link below. 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 WITHDRAWN AND REPLACED — the entry was wrong, and a corrected version of it appears in the entries below.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 An entry we have WITHDRAWN is removed from this page and kept in the record, with its reason.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.| relation to other work | claim_relations | how this paper stands to another: agrees, contradicts, extends, qualifies |
| claim type | claim_type | the shape of the claim: existence, mechanism, method, critique, prediction, null |
| concept edge | concept_edges | an asserted relation between two concepts, with the sentence asserting it |
| concept | concepts | an idea the paper works with, quoted where the paper uses it |
| content tag | content_tags | a finding, definition, quantity, statistic or variable named in the paper |
| contribution | contribution | what the paper claims to add |
| control | controls | an experiment run to eliminate a rival explanation |
| limitation | limitations | what the paper says its evidence cannot support |
| mechanism | mechanisms | how the reported effect is said to come about |
| method | method | how the paper produced its evidence |
| object of study | object_of_study | what the paper is about — the thing it investigates |
| phenomenon | phenomenon | the effect or regularity under study |
| research question | research_question | the question the paper sets itself |
| sense | senses | a distinct use of a term, quoted where the paper fixes it |
| transfer | transfers | a method or concept carried in from another field |
Relations on this page carry a quotation for the claim. Where a relation was separately evidenced — and so tested against the vocabulary for its type — the row says so; 2 of 2 relations here carry that mark. The rest are the tagger’s reading of the sentence quoted beside them.
| facet | tag | field | conf. | whose words |
|---|---|---|---|---|
| object_of_study | 22 AI ethics guidelines - which issues each one covers, which it omits, and whether any of it changes what practitioners do Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | object_of_study | the paper's words | |
| research_question | whether ethical guidelines have an actual impact on human decision-making in AI and machine learning Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | research_question | the paper's words | |
| contribution | a comparison of 22 guidelines that tallies coverage AND absence, sorts the absences into justified and unjustified, and states that the selection method decided the list Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | contribution | the paper's words | |
| claim_type | critique Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | claim_type | the paper's words | |
| phenomenon | a normative vocabulary assembled by many bodies at once, whose commonest terms turn out to be the ones with engineering solutions Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | phenomenon | the paper's words | |
| method | review Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: review | method | the paper's words | |
| mechanisms | accountability, privacy and fairness appear in about 80% of the guidelines, and the author's observation is that these are exactly the aspects for which technical fixes exist or are being built. A NORMATIVE VOCABULARY SELECTED BY TRACTABILITY: what can be engineered gets written down, and what cannot - political abuse, ecological cost, democratic erosion - does not Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | mechanisms[0] | 0.95 | the tagger's words |
| mechanisms | excluding robot ethics guidelines means the list of issues lacks robotics issues. The author states this BEFORE reporting the omissions, so every absence in the paper is already qualified by the selection that produced it - which is the discipline a negative finding needs and almost never gets Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: the sample decides the absence list | mechanisms[1] | 0.95 | the tagger's words |
| mechanisms | an issue can be missing without any good reason - political abuse, hidden social and ecological costs - or missing for a reason that can be stated. The author refuses to count them together, which turns a list of gaps into two different claims Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: two kinds of absence | mechanisms[2] | 0.95 | the tagger's words |
| mechanisms | a marker goes in only where an issue is explicitly discussed in one or more paragraphs - unless the guideline is so short that it is nothing but brief mentions, in which case brief mentions count. THE EXCEPTION IS WHERE THE MEASUREMENT WOULD OTHERWISE PENALISE BREVITY RATHER THAN SILENCE Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | mechanisms[3] | 0.95 | the tagger's words |
| mechanisms | industry guidelines suggest to legislators that self-governance suffices and no specific laws are needed. So the existence of an ethical vocabulary is itself an argument against a binding one - the guidelines are not only a record of values, they are an intervention in whether values become rules Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: ethics as a substitute for law | mechanisms[4] | 0.95 | the tagger's words |
| limitations | the sampling method has an effect on the list of issues and omissions - excluding robot ethics guidelines removes robotics issues from the result Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | limitations[0] | 0.95 | the tagger's words |
| limitations | the selection is biased towards western and northern documents and excludes guidelines not written in English Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | limitations[1] | 0.95 | the tagger's words |
| limitations | the evaluation is semi-systematic rather than a full analysis of every available soft-law or non-legal norm document Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | limitations[2] | 0.95 | the tagger's words |
| limitations | documents about big data, algorithms or robotics, and papers on single aspects of AI ethics, were excluded Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: PAGE FURNITURE IS SPLICED INTO THE 'evidence' QUOTE - the publisher's page mark, a page number and the author running head: '1 3 102 T. Hagendorff'. The words on either side of it were never continuous on the page; read the quote with it deleted. | limitations[3] | 0.95 | the tagger's words |
| limitations | the overview of missing issues is stated as non-exhaustive Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | limitations[4] | 0.95 | the tagger's words |
| transfers | systematic-review methodology from evidence-based medicine, applied semi-systematically to normative documents rather than to trials Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | transfers[0] | 0.95 | the tagger's words |
| controls | a stated marking criterion - explicit discussion in one or more paragraphs - applied before any tally, with the exception for guidelines that are entirely brief mentions Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. eliminates: a word-count masquerading as coverage; a guideline that names privacy once does not score the same as one that discusses it · outcome: held | controls[0] | 0.95 | the tagger's words |
| controls | the effect of the sampling method on the omissions list, stated in the same section as the omissions Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. eliminates: an absence being read as a property of the field when it is a property of the sample - the robotics example is given as the worked case · outcome: held | controls[1] | 0.95 | the tagger's words |
| controls | omissions sorted into those without any good reason and those that can be justified, rather than counted as one figure Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. eliminates: a gap count in which a deliberate scope decision and a genuine blind spot are the same number · outcome: held | controls[2] | 0.95 | the tagger's words |
| controls | selection criterion declared as the discernible intention of comprehensive mapping rather than depth of detail Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. eliminates: the longest documents dominating the sample, which would make coverage a function of document length · outcome: held | controls[3] | 0.95 | the tagger's words |
| controls | the author's own selection bias towards western, northern and English-language documents, named as a shortcoming Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. eliminates: guidelines from outside the western, northern, English-language world - the author names the bias in his own corpus and does not correct it · outcome (stated by the tagger, no quotation): absent-declared-unreported | controls[4] | 0.95 | the tagger's words |
| concepts | omissions Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: omissions | concepts[0] | 0.95 | the paper's words |
| concepts | sampling method Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: sampling method | concepts[1] | 0.95 | the paper's words |
| concepts | political abuse Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: political abuse | concepts[2] | 0.95 | the paper's words |
| concepts | technical fixes Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: technical fixes | concepts[3] | 0.95 | the paper's words |
| concepts | accountability Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: accountability | concepts[4] | 0.95 | the paper's words |
| concepts | green markers Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: green markers | concepts[5] | 0.95 | the paper's words |
| concepts | Isolated mentions Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: Isolated mentions | concepts[6] | 0.95 | the paper's words |
| concepts | AI ethics Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: PAGE FURNITURE IS SPLICED INTO THE 'evidence' QUOTE - the publisher's page mark, a page number and the author running head: '1 3 102 T. Hagendorff'. The words on either side of it were never continuous on the page; read the quote with it deleted. also in: AI ethics | concepts[7] | 0.95 | the paper's words |
| concepts | big data Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: PAGE FURNITURE IS SPLICED INTO THE 'evidence' QUOTE - the publisher's page mark, a page number and the author running head: '1 3 102 T. Hagendorff'. The words on either side of it were never continuous on the page; read the quote with it deleted. also in: big data | concepts[8] | 0.95 | the paper's words |
| concepts | depth of detail Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: depth of detail | concepts[9] | 0.95 | the paper's words |
| concepts | machine consciousness Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: PAGE FURNITURE IS SPLICED INTO THE 'evidence' QUOTE - the publisher's page mark, the title running head and a page number: '1 3 The Ethics of AI Ethics: An Evaluation of Guidelines 105'. The words on either side of it were never continuous on the page; read the quote with it deleted. also in: machine consciousness | concepts[10] | 0.95 | the paper's words |
| senses | an omission - UNJUSTIFIED: missing or severely underrepresented without any good reason, such as political abuse or the hidden social and ecological costs of AI Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | senses[0] | 0.95 | the tagger's words |
| senses | an omission - JUSTIFIED: absent for a reason the analysis can state, which the author separates out rather than counting alongside the first Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | senses[1] | 0.95 | the tagger's words |
| senses | an issue covered - AS CODED: explicitly discussed in one or more paragraphs, with isolated mentions not counted Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | senses[2] | 0.95 | the tagger's words |
| senses | an issue covered - THE EXCEPTION TO THE CODING: unless the guideline is so short that it consists entirely of brief mentions, in which case brief mentions count Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | senses[3] | 0.95 | the tagger's words |
| senses | `AI` - AS USED IN THE GUIDELINES: the only term they use, referred to exclusively and almost never replaced by more specific terminology Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | senses[4] | 0.95 | the tagger's words |
| senses | `AI` - AS THE AUTHOR READS IT: a collective term for a wide range of technologies, or an abstract large-scale phenomenon - which is why no guideline goes into technical detail Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | senses[5] | 0.95 | the tagger's words |
| claim_relations | critiques → the AI ethics guidelines themselves, on the grounds that they lack any mechanism to reinforce their own normative claims Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | claim_relations[0] | 0.95 | the tagger's words |
| claim_relations | critiques → industry self-governance, as a device for suggesting to legislators that no specific laws are necessary Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | claim_relations[1] | 0.95 | the tagger's words |
| claim_relations | critiques → the guidelines' exclusive use of the term AI, which is a collective term for a wide range of technologies and lets every document avoid technical detail Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | claim_relations[2] | 0.95 | the tagger's words |
| claim_relations | qualifies → its own omissions list, by stating that the sampling method produced it and that the selection is western, northern and English-language Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | claim_relations[3] | 0.95 | the tagger's words |
| concept_edges | green markers —distinct-from→ Isolated mentions Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. differs in: whether the issue was discussed or merely named · the relation itself is separately evidenced · strength of the relation: 0.95 · differs in (stated by the tagger, no quotation): whether the issue was discussed or merely named · the relation stated: an issue counts as covered only where it is explicitly discussed in one or more paragraphs; a bare mention does not - unless the whole guideline is bare mentions
Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the span explains which issues got a green marker; distinct-from is loose. also in: green markers · Isolated mentions | concept_edges[0] | 0.95 | the tagger's words |
| concept_edges | AI ethics —distinct-from→ big data Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. differs in: which technology the document is about, which is the boundary the selection drew · the relation itself is separately evidenced · strength of the relation: 0.95 · differs in (stated by the tagger, no quotation): which technology the document is about, which is the boundary the selection drew · about this quote: PAGE FURNITURE IS SPLICED INTO THE 'evidence' QUOTE - the publisher's page mark, a page number and the author running head: '1 3 102 T. Hagendorff'. The words on either side of it were never continuous on the page; read the quote with it deleted. PAGE FURNITURE IS SPLICED INTO THE relation evidence QUOTE - the publisher's page mark, a page number and the author running head: '1 3 102 T. Hagendorff'. The words on either side of it were never continuous on the page; read the quote with it deleted. · the relation stated: documents about big data, algorithms or robotics were excluded from the sample even where they are ethics documents, so the issues peculiar to them are absent from the list by construction | concept_edges[1] | 0.95 | the tagger's words |
| content_tags | the most frequently mentioned ethical aspects are exactly those for which technical fixes exist or are being developed Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | content_tags.findings[0] | 0.95 | the tagger's words |
| content_tags | the sampling method used to select the guidelines has an effect on the list of issues and omissions Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | content_tags.findings[1] | 0.95 | the tagger's words |
| content_tags | not all omissions can be treated equally - some are missing without any good reason and some can be justified Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | content_tags.findings[2] | 0.95 | the tagger's words |
| content_tags | hardly any guideline discusses political abuse of AI - propaganda, bots, fake news, deepfakes, micro targeting, election fraud Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | content_tags.findings[3] | 0.95 | the tagger's words |
| content_tags | machine consciousness is omitted from all examined guidelines Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: PAGE FURNITURE IS SPLICED INTO THE 'evidence' QUOTE - the publisher's page mark, the title running head and a page number: '1 3 The Ethics of AI Ethics: An Evaluation of Guidelines 105'. The words on either side of it were never continuous on the page; read the quote with it deleted. | content_tags.findings[4] | 0.95 | the tagger's words |
| content_tags | Only the AI Now 2019 Report notes that AI producer and consumer practices may themselves contradict sustainability goals. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | content_tags.findings[5] | 0.95 | the tagger's words |
| content_tags | the guidelines refer exclusively to the term AI and never or very seldom use more specific terminology Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | content_tags.findings[6] | 0.95 | the tagger's words |
| content_tags | AI ethics lacks mechanisms to reinforce its own normative claims, and deviations from the codes have no consequences Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | content_tags.findings[7] | 0.95 | the tagger's words |
| content_tags | industry ethics guidelines serve to suggest to legislators that self-governance is sufficient and no specific laws are necessary Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | content_tags.findings[8] | 0.95 | the tagger's words |
| content_tags | issues absent from the guidelines, divided into those missing or severely underrepresented without any good reason and those that can be justified Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: omissions | content_tags.definitions[0] | 0.95 | the tagger's words |
| content_tags | inserted only where an issue is explicitly discussed in one or more paragraphs; isolated mentions without further explanation do not count, unless the guideline consists entirely of brief mentions Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: green markers | content_tags.definitions[1] | 0.95 | the tagger's words |
| content_tags | AI ethics guidelines analysed and compared Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | content_tags.quantities[0] | 0.95 | the tagger's words |
| content_tags | of all guidelines mention accountability, privacy or fairness - the three aspects with technical fixes Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. | content_tags.quantities[1] | 0.95 | the tagger's words |
| content_tags | omissions Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: omissions | content_tags.variables[0] | 0.95 | the tagger's words |
| content_tags | sampling method Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: sampling method | content_tags.variables[1] | 0.95 | the tagger's words |
Tags that could not be placed in the text:
What was deliberately not tagged, and why, is recorded per paper by the tagger. Where a paper has no such record, this page says so rather than implying that nothing was declined.