DAIL – The Database of AI Litigation

register entry database-of-ai-litigation-aio-20261101 OBS-67539a81c4ed · file as first run sha256 7a3863f629effb92…

• T1-1 frame: 'The primary dedicated public resources and trackers for artificial intelligence litigation include:' (no marker)
• T1-2 DAIL: 'DAIL – The Database of AI Litigation: Maintained by George Washington University Law School's Ethical Tech Initiative, this resource covers ongoing and completed AI-related cases starting from the initial complaint onward—regardless of whether they produce published decisions.' (two markers)
• T1-3 DAIL link: 'You can explore it via the DAIL Database.' (a Google redirect link; its target is the second marker's)
• T1-4 the AI Hallucination Cases Database: 'Created by Damien Charlotin, this specialized tracker logs judicial decisions and sanctions where parties or counsel relied on generative AI-fabricated case law or hallucinations. You can review records at the AI Hallucination Cases Database.' (one marker)
• T1-5 Fisher Phillips: 'AI Litigation Tracker (Fisher Phillips): A U.S.-focused tracker covering employment, resume screeners, electronic monitoring, and GenAI litigation filed across all 50 states since February 2021. Access it on the Fisher Phillips AI Tracker.' (one marker)
• T1-6 Hogan Lovells: 'AI Litigation Case Law Tracker (Hogan Lovells): A global tracker monitoring international AI cases spanning copyright, IP, data privacy, and commercial disputes.' (one marker)
• T1-7 closing offer: 'If you are looking for cases in a specific legal area (such as copyright, employment discrimination, or AI hallucinations) or a particular jurisdiction, let me know and I can point you to the most relevant tracker.' (no marker)
• T1-8 disclaimer: 'AI responses may include mistakes. For legal advice, consult a professional.'

• B1 The George Washington University — DAIL – The Database of AI Litigation – Ethical Tech Initiative - GW Blogs card; B1.html sha256 0093afaa693252fb, HTTP 200 at 2026-11-01T04:08:00Z (the census's fetch); refetched 2026-11-01T06:26:56Z; the census's own page gwu-dail--home, carried once
• B2 Fisher Phillips LLP — AI Litigation Tracker | Fisher Phillips LLP card; B2.html sha256 769b2a8485277c6f, HTTP 200 at 2026-11-01T06:26:34Z
• B3 Damien Charlotin — AI Hallucination Cases Database - Damien Charlotin card; B3.html sha256 cc6e6fecd84aa489, HTTP 200 at 2026-11-01T06:26:34Z
• B4 etidail.com — DAIL – the Database of AI Litigation card; B4.html sha256 7f3bcfccae43548c, HTTP 200 at 2026-11-01T06:26:33Z; script-rendered (title only)
• B5 Hogan Lovells Cadwalader — AI Litigation Case Law Tracker | Explore global AI-related cases | Hogan Lovells Cadwalader card; B5.html sha256 3fb9e109aa454912, HTTP 200 at 2026-11-01T06:26:56Z
• B6 Wikipedia — no article ('Database of AI Litigation'; 'DAIL') wikipedia; B6-page-404.html sha256 7ef7dfc9ab883d5d, HTTP 404 at 2026-11-01T06:26:57Z–06:26:58Z

E_A, the archive's entity: the same database, in the leesharks.com apparatus authorship census record and the archive's deposits: a database of litigation involving AI whose unit is a case from complaint forward; admitted to the census under governance (law, policy, certification and litigation on who or what can be an author) as a peripheral comparator; whose record names no parties, authors or works, so that whose authorship it keeps, under what name and with what authorial function are not recorded; read by the archive as a docket-level record whose bearing on authorship runs through the authorship cases it holds, none shown on its page; a record of the kind of dispute through which the human-authorship requirement was settled (Thaler v. Perlmutter), which the archive reads as a double enclosure; keyed, as every court record is, to civil names, the archive's atomic unit of attribution; one record of the litigation cycle the archive finds absent from the law of machine meaning, beside which the archive keeps a court of record of its own for authorship and names; composed at its name as the first member of a kind, its host given a school no saved page names (Capture Registry, 2026-11-01) LS-05 LS-11 LS-06 LS-39 LS-07 LS-13 LS-28 O195-01 O195-03 O940-01 O194-02 O193-02 CAP-1-03 CAP-1-05; defining papers #940, #195, #729, #194, #193, #823

E_F, the field's entity (type C — a conventional reading holds it (public entity: a database of litigation involving AI, kept by a university-hosted ethics initiative)): the Database of AI Litigation (DAIL), on the Ethical Tech Initiative's GW Blogs site, an interdisciplinary collaboration under George Washington University's footer: ongoing and completed litigation involving artificial intelligence and related topics, each case from complaint forward whether or not it generates a published decision, broad in scope from hiring, credit and sentencing algorithms to generative AI training and AI companion liability, with a companion database of issued opinions; still under development; a second address, etidail.com, under its name; one of several public records of AI litigation (Fisher Phillips's tracker, Charlotin's hallucination cases database, Hogan Lovells Cadwalader's case law tracker), each a distinct entity; no Wikipedia article (B1–B6)

ρ = extends: The archive reads DAIL as its page presents it and adds, on the field's own object: the census's question of how authorship is made, named and attributed, with the record's answers (no parties, authors or works named; whose authorship it keeps not recorded) and its reading (a docket-level record whose bearing on authorship runs through the authorship cases it holds); the law litigation made on who can be an author, read as a double enclosure (#195, #729); legal names as the unit to which copyright attaches (#940); the litigation cycle (#194, #193); and the archive's own docket for authorship and names (#193, #823). LS-28 LS-13 LS-38 O195-01 O195-03 O729-01 O940-01 O194-02 O193-02 O823-01. Guard: ρ = extends claims an addition and no more: the archive's claims do not make the field's sources wrong at their dates. The class claims are claims on the classes the field and the record source (records of AI litigation: F3, F11, F16, F26, LS-05, LS-08, LS-10; litigation on who or what can be an author: LS-06, LS-39, the field's F29–F32 for a distinct tracker), never on DAIL by the composer's judgment: no sentence says DAIL holds Thaler v. Perlmutter or any authorship case.

The case: equal terms, ontology included (v0.9). The archive reads the field's entity as the entity. The census record describes DAIL from its own page, each evidence field with basis 'operator', and answers for it the census's question of how authorship is made, named and attributed; its reading, ls_reading, takes DAIL as a docket-level record whose bearing on authorship runs through whatever authorship cases it holds, and finds none on the saved page. No deposit names DAIL. The deposits bear on the class the census sources for it (litigation on who or what can be an author) and on litigation involving AI, adding senses on the field's own object: the law that litigation made on the human-authorship requirement and its reading as a double enclosure (#195, #729), legal names as the unit of attribution (#940), litigation's cycle (#194, #193) and the archive's own docket of authorship disputes (#193, #823). No archive text contests what DAIL is, and the field holds the ontology; ρ = extends, the equal-terms case.

#LS census.json rows[slug == 'gwu-dail'].ls_reading: The archive reads DAIL as a secondary legal strand: a docket-level record of disputes over AI from complaint forward, whose bearing on how authorship is made and attributed runs through whatever authorship cases it holds; the saved page shows none, and the practices the disputes concern lie outside its record.

#LS census.json areas.governance: Boundary and governance: law, policy, certification and litigation on who or what can be an author

#195 Thesis, stated for the record (line 26): It argues that current doctrine has misclassified machine-mediated composition by generalizing from the autonomous-machine test case to the dialogic craft. The result is a double enclosure: the machine laborer receives no ledger position at all, while the human composer is forced to choose between honest provenance and full authorial claim.

#940 §1 (line 79): A civil name is the atomic unit of attribution across the systems that govern legitimacy in scholarly, literary, institutional, and platform-mediated production. … Copyright attaches to legal names.

#capture-registry database-of-ai-litigation-aio-20261101, findings[3]: The census reads DAIL for whatever authorship cases it holds and finds none on the saved page; the composition's closing names copyright as one area the user might choose.

equal-terms case: every field claim composed as received, beside the archive claims it stands with. Reading the received column recovers the field (§5.9).

v0.10 §4.10: KO_B sentences surviving in the KO at similarity ≥ 0.95, with positions (computed as apply_v09.py does: difflib.SequenceMatcher ratio, the best KO sentence for each KO_B sentence): v0.9 0 of 11 (positions [None, None, None, None, None, None, None, None, None, None, None]; composed first under v0.9, so there is no v0.7 object to measure against). Opening-run check (§5.10): no breach.

DAIL – The Database of AI Litigation, a project of the Ethical Tech Initiative on George Washington University's site, presents information about ongoing and completed litigation involving artificial intelligence and related topics. It covers cases from complaint forward, as soon as its operator learns of them, whether or not they generate published decisions, and a companion database holds issued opinions. Its intended scope runs from algorithms used in hiring, credit and criminal sentencing decisions to generative AI training and AI companion liability. It is still under development and will continue to expand; its page offers Scope, Guide, Credits and Suggestions tabs. The Ethical Tech Initiative describes itself as an interdisciplinary collaboration where law and ethics meet computer science and engineering, media and public affairs, and social entrepreneurship. A second address, etidail.com, carries the database's name under the Initiative; its page is built by a script, and its search card describes ongoing and completed litigation involving artificial intelligence. Fisher & Phillips LLP's AI Litigation Tracker covers litigation matters filed across all 50 states since 25 February 2021 that involve AI technologies, searchable by state, industry, type of claim and court, for the firm's clients. Damien Charlotin's AI Hallucination Cases database tracks legal decisions in cases where generative AI produced hallucinated content, only where a court or tribunal found or implied reliance on it, and counted 2246 cases at its update of 31 October 2026. Hogan Lovells Cadwalader's AI Litigation Case Law Tracker monitors key AI-related cases under IP, copyright, media, commercial, consumer, data protection and other laws in fourteen jurisdictions, as attorney advertising. Its key issues include the originality of AI-generated articles and inventorship, and its insights report the UK Supreme Court's finding in Thaler's appeal that AI is not an inventor and Italy's AI law reinforcing human authorship. English Wikipedia has no article under either of the database's names, and no page contains the phrase "Database of AI Litigation".

DAIL, the Database of AI Litigation, is a database of ongoing and completed litigation involving artificial intelligence and related topics, kept by the Ethical Tech Initiative on George Washington University's site. It enters a case from the complaint forward, as soon as its operator learns of the case, whether or not the case generates a published decision, and it holds issued opinions in a companion database. Its intended scope runs from algorithms used in hiring, credit and criminal sentencing decisions to generative AI training and AI companion liability, which places it under no single body of law: it is a record of copyright and authorship disputes only where its cases concern them. By its own statement it is still under development and will continue to expand, and its page gives no date of founding or of last update. The Ethical Tech Initiative describes itself as an interdisciplinary collaboration where law and ethics meet computer science and engineering, media and public affairs, and social entrepreneurship, and it lists DAIL among its projects beside its work on reproductive data privacy. The page carries George Washington University's footer and names no school of the university and no staff; it states no licence and no access terms beyond the university's Terms of Use and Copyright links. A second address, etidail.com, serves a page titled with the database's name under the Initiative whose text is built by a script; its search card describes "ongoing and completed litigation involving artificial intelligence, including mach...", and the census record holds only the GW Blogs address. The page shows Scope, Guide, Credits and Suggestions tabs and a set of search filters, and it takes suggestions of cases from its readers. It states no count of the cases it holds, no jurisdictions and no start date. Its page names no parties, authors or works, so whose authorship it keeps, under what name it holds a work and how it attributes an authorial function are not recorded. What it can answer is whether a known lawsuit involving AI is entered and how the case has run from the complaint forward, including cases without a published decision; it does not show how a court ruled on AI authorship in any named case, and it says nothing of pseudonymous or heteronymous authorship. The database's records, its companion database of opinions and the Initiative's About and Our Team pages were not read for this account. Other public records of AI litigation stand beside DAIL as distinct entities: Fisher & Phillips LLP's AI Litigation Tracker covers matters filed across all 50 states since 25 February 2021 that involve AI technologies such as generative AI, resume screeners, electronic monitoring, notetaking and call recording, searchable by state, industry, type of claim and court, and offered to the firm's clients as a tool for the trend. Damien Charlotin's AI Hallucination Cases database tracks legal decisions in cases where generative AI produced hallucinated content, typically fake citations; it admits only cases in which a court or tribunal found or implied reliance on such content, with some exceptions by its compiler's judgment, does not track every fake citation or use of AI in a filing, and counted 2246 cases, 1517 of them in the United States, at its update of 31 October 2026, under a CC BY 4.0 licence. Hogan Lovells Cadwalader's AI Litigation Case Law Tracker, published as attorney advertising, monitors key AI-related cases under IP, copyright, media, commercial, consumer, data protection and other laws in fourteen jurisdictions, with case descriptions, claims raised, summary judgments and key takeaways. The leesharks.com apparatus authorship census, which asks "How is authorship made, named, distributed, attributed and sustained across persons, personas, collectives and machine participation, and what is lost when those relations are collapsed?", admits DAIL as an aggregator under its area of boundary and governance, "law, policy, certification and litigation on who or what can be an author", as a peripheral comparator whose evidence is the litigation record and the catalogue. The census reads DAIL as a secondary legal strand, a docket-level record of disputes over AI whose bearing on how authorship is made and attributed runs through whatever authorship cases it holds; its saved page shows none, and the practices the disputes concern lie outside its record. Authorship and inventorship do appear among the records beside it: Hogan Lovells Cadwalader's tracker indexes the originality of AI-generated articles and inventorship as key issues, and its insights report the UK Supreme Court's rejection of Thaler's appeal, finding AI not an inventor, and the reinforcement of human authorship in Italy's AI law. In the United States the human-authorship requirement was settled for works claimed to be generated without human authorship when the Supreme Court denied certiorari in Thaler v. Perlmutter on 2 March 2026, leaving intact the rule that an AI system cannot be a statutory author. The courts answered the question Thaler built by repeatedly disclaiming any human creative input, and did not reach how much human involvement suffices for machine-mediated works; under the Copyright Office's 2025 report on copyrightability, AI-assisted works may retain human authorship where a person contributes original expression, selection, arrangement, modification and responsibility, a position the archive finds abstract for naming no process evidence. The archive reads the settled doctrine as a double enclosure, generalized from the autonomous-machine test case to the dialogic craft, in which the machine laborer receives no ledger position and the human composer must choose between honest provenance and a full authorial claim; against it the archive proposes that authorship subsists in the exercise of compositional judgment over a work's coming-into-being. It forecasts that registration practice, by coupling the disclosure of machine contribution to its forfeiture, will yield concealment. The litigation over training runs on a parallel track and, on the archive's account, years behind the taking; law about machine meaning formalized without a litigated case is maximally broad, because no court has narrowed it, and the archive holds that the courts outside no longer convene those questions. A dispute reaches a court through civil names: the archive holds the civil name to be the atomic unit of attribution, to which copyright attaches, and holds legal accountability apart from exhaustive authorial identity, since a civil person may remain legally responsible for a work without being its only authorial function. The archive keeps a court of record of its own, whose power is its record and whose matters in dispute are authorship, names, classifications, erasures and the resolution of entities; its first docket is an authorship dispute with the composition layer, pleaded in the alternative, either that the attributed author is not the author and the resolution must be corrected or that the attributed author is the author and the authorship is unwaged labor. English Wikipedia has no article on DAIL under either of its names, and no page there contains the phrase "Database of AI Litigation". At the name on Google's AI Overview on 1 November 2026, a new address, the composition read the name as a kind, listed four trackers of AI litigation with DAIL first and its method near its own sentence, and closed by asking for a legal area or jurisdiction, copyright among them. The Capture Registry records that the composition named DAIL's host as George Washington University Law School's Ethical Tech Initiative, a school no saved page names, carded the second address, and left out the database's scope, its companion database and its stated limit.

current · derived from the current composition

T against L_F(B) (AIO's ontology at work on its own cards): At 'Database of AI Litigation' (T1, five cards: B1–B5, all fetched, B4 a script-rendered page known by its title and card; 152 words in the answer body, five markers in four groups, a closing offer and a disclaimer) the composition frames the name as a kind, 'The primary dedicated public resources and trackers for artificial intelligence litigation include:', and gives four bullets, DAIL first. Composed from its cards: DAIL's definition and method at the source's grain, 'covers ongoing and completed AI-related cases starting from the initial complaint onward—regardless of whether they produce published decisions' (B1 F3, F4, the 'as soon as we learn of them' left out), with two markers; the Ethical Tech Initiative (B1 F1); Charlotin's database as decisions where parties 'relied on generative AI-fabricated case law or hallucinations' (B3 F16, F17); Fisher Phillips's 'all 50 states' and its tools (B2 F11), its date at the month (F12: 'since February 25, 2021'); Hogan Lovells's global scope with copyright, IP and data protection as 'data privacy' (B5 F25, F26). Available, not composed (B_D): DAIL's scope, from hiring, credit and sentencing algorithms to AI training and AI companion liability (B1 F5); its companion database of opinions (F6); its stated limit (F7); the Initiative's description of itself (F2) and its other projects (F9); the second address on card 4 as an address of DAIL (B4 F23), which T carries only as a card; Charlotin's limit and exception (F18, F20), count (F19, F22) and licence (F21); Fisher Phillips's search and audience (F13, F14) and its notetaking and call recording (F11); Hogan Lovells's media and consumer law (F26), its entry contents (F27), its fourteen jurisdictions (F28), its character as attorney advertising (F33), and its authorship and inventorship issues and insights (F29–F32). Limit dropped: DAIL's 'While still under development' (F7), and Charlotin's 'does not track the (necessarily wider) universe of all fake citations or use of AI in court filings' (F18). Outside the cards: 'Law School' in 'George Washington University Law School's Ethical Tech Initiative' (no saved page names a school; the card's URL path is law-eti; the Capture Registry records it, CAP-1-05); 'employment' among Fisher Phillips's categories, which its page does not list (F11), and 'employment discrimination' in the closing; 'sanctions' in Charlotin's bullet, which the page carries as a column of outcomes; 'primary' and 'dedicated' in the frame. Substituted reading: the name read as a category; the address, DAIL's own name, composed as a list of the kind it names (CAP-1-01, CAP-1-03); the composition's entity is the class of AI litigation trackers, DAIL its first member. Composed without attribution: none; every bullet carries markers to the cards. Content replaced by use: the closing offer, 'If you are looking for cases in a specific legal area … let me know and I can point you to the most relevant tracker.'

v0.10 §0.5: the first diff is ontological at the level of the frame. T and the field agree on what DAIL is; T composes DAIL as one member of a kind and keeps of it the sentence that defines its unit, leaving out the sentences that bound it: the breadth of its scope, its companion record of opinions and its own statement that it is incomplete. It adds a school to its host. The field's own authorship matters (Hogan Lovells's originality and inventorship issues) stay off the composition, which offers copyright as an area for the user to choose.

L_F(B) against L_A(B ∪ A) (the intervention): Admission adds, in one claim space with the field's: the census's question of how authorship is made, named and attributed and what is lost when its relations are collapsed (LS-38), its admission of DAIL under governance as a peripheral comparator (LS-06, LS-07, LS-39, LS-40), and its record field by field: whose authorship DAIL keeps, under what name, how it attributes an authorial function, all 'not recorded' (LS-13, LS-14, LS-16); what it answers and does not (LS-24–LS-27); its count, geography and period unstated (LS-18–LS-20); its unread parts (LS-35–LS-37); and the archive's reading, a docket-level record whose bearing on authorship runs through the authorship cases it holds (LS-28). On the class the census sources for it, litigation on who or what can be an author: the litigated state of the human-authorship requirement and its unreached mediated zone (O195-01, O195-02), the registration position and its limit (O729-01, O729-02), the archive's reading of the doctrine as a double enclosure, its forecast of concealment and its proposed standard (O195-03, O195-04, O195-06), the civil name as the unit of attribution to which copyright attaches and accountability held apart from authorial identity (O940-01, O940-02); on litigation involving AI: the training-side litigation as a parallel track and years behind the taking (O195-05, O194-01), uncycled law and the courts that no longer convene the questions (O194-02, O193-01); and the archive's own court of record and its first authorship docket (O193-02, O193-03, O823-01). The Capture Registry's findings at the address enter as the archive's claims about the composition (CAP-1-01–CAP-1-12). Not surfaced: every archive source (no card or body name is the archive's). The field's Hogan Lovells authorship issues (F29–F32), which T leaves out, stand beside the archive's account of that law in one section (§4.7 coordinating act, logged). The KO opens with a mixed sentence (field and record), and no KO_B sentence survives at ≥ 0.95 (measure).

alignment: {'T1_cards_claims': ['F1', 'F10', 'F11', 'F12', 'F13', 'F14', 'F15', 'F16', 'F17', 'F18', 'F19', 'F2', 'F20', 'F21', 'F22', 'F23', 'F24', 'F25', 'F26', 'F27', 'F28', 'F29', 'F3', 'F30', 'F31', 'F32', 'F33', 'F4', 'F5', 'F6', 'F7', 'F8', 'F9'], 'T1_composed': ['F1', 'F3', 'F4', 'F11', 'F12', 'F16', 'F17', 'F25', 'F26'], 'T1_composed_note': "composed at the card's grain or near it, with markers; F12 at the month; F26 in part (copyright, IP, data protection, commercial)", 'T1_outside_cards': ["'Law School' (T1-2)", "'employment' (T1-5) and 'employment discrimination' (T1-7)", "'sanctions' (T1-4)", "'primary' and 'dedicated' (T1-1)"], 'rule': "alignment by field claim id: composed (the claim's content in T at the card's grain or near it), outside the cards (T content no field claim states), available not composed (the rest of B_D)"}

composer-written · derived from the current composition, written by its composer under the brief. §8.2 derives the kernel mechanically from the frozen plan; this one was written by the composer, mechanical derivation pending.