provenanceerasure.org ranks first. Defines provenance erasure as 'the systematic removal, loss, or intentional severing of the origins, lineage, or authorship of an idea, digital asset, or physical artifact.' 9+ sources cited including Institute of Historic Research. Zenodo shows Provenance After AI with the PER formula. Author name absent from Overview text.

Full AI Mode transcript (17 June 2026): The Provenance Erasure Rate (PER) is an evaluation metric used to measure attribution loss in artificial intelligence and synthetic composition systems. It calculates the fraction of human-authored or source facts that a system's retrieval or generation process dissolves, omits, or fails to properly attribute when composing an answer. Core Metric: PER quantifies the magnitude of provenance loss within AI generation, summarization, or synthesis tools. Provisional Formula: PER = 1 - (retained provenance units / total attributable source claims) Erasure Skew (Ω): A related measurement that evaluates whether attribution loss falls evenly across all sources, or systematically strips out low-power/small creators while favoring high-power sources.

provenanceerasure.org — Lee Sharks ranks first organic. Zenodo shows Provenance After AI with PER formula visible: PER = 1 − (retained provenance units / ...). Framework retrievable by name. Author retrievable by site. Overview dissolves the connection.

THE TERM HAS THREE LEGITIMATE HOMES AND THE ARCHIVE IS THE FOURTH. Card 2 is ACM on SOCIETAL PROVENANCE as a grounding framework for archival description — the same word, the same discipline, an adjacent argument. Card 3 is on GDPR: "regulations require systems to support data erasure but LEAVE THE DEFINITION OF ERASURE OPEN TO INTERPRETATION." A legal-compliance paper about the ambiguity of the word the archive uses as a metric.

Card 1 is the sharpest and is not the archive’s: "There is a growing pattern in this field: A BUILDER PUBLISHES A SYSTEM. OTHERS ENTER THE…". An independent scholar describing the same dynamic of uncredited adoption that the archive documents from the inside.

provenanceerasure.org appears as card 4 — the archive holding a domain named for the term, present in the basin, ranked beneath three unrelated treatments of it. Not erasure and not collision: a genuine field with several occupants, in which the archive is one voice and not the loudest.

**Longitudinal with `provenance-erasure-1` and `provenance-erasure-2` (13 June) and `provenance-erasure-rate-adoption` (17 June).** June recorded the definition with the author’s name absent from the Overview text. Tonight the term operates as general vocabulary: as screenshotted, *the loss or removal of an item’s history, origin, or authorial lineage* through *artificial intelligence synthesis, data compression, or institutional actions like colonial bureaucracy and museum record-keeping*, cited to Oxford Academic +5; as transcribed after regeneration, a three-context taxonomy (digital/AI systems; institutional and museum archives — the British Toshakhana; GDPR dependent-inference erasure). The authorial credit has migrated from the composed layer to the organic layer, where it is now definitional: provenanceerasure.org first, snippet reading *Defined by Lee Sharks (2026) within...*, and the alexanarch Capture Registry manuscript record carrying the measurement programme — *PER (provenance erasure rate), Erasure Skew, and atomic-token preserva...* Third-party circulation is present and dated: Terry Snyder’s LinkedIn post (2 August) — *PROVENANCE ERASURE IS BECOMING A BUSINESS MODEL. There is a growing pattern in this field: A builder publishes a system. Others enter the ...* — the term doing independent work in professional discourse nine days before this capture. Injection at the floor: the layer inserts an Erasure (the band) music video, *Where In The World* — lexical adjacency again. A coined term whose author is no longer needed in the definition because the definition’s first source is the author’s own surface: mantle consolidation of vocabulary, seen from outside.