[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82570-en":3,"doc-seo-82570-105":29,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},82570,962075006959,"Anda","https://ap-avatar.wpscdn.com/avatar/e0002397efbe92a78e?_k=1776741047341049297",8,"Research & Report","From Runtime Records to Legal Findings: An Evidentiary-Adequacy Criterion for Agentic AI Oversight","Agentic AI systems produce runtime records, logs, traces, and audit artefacts, yet such artefacts alone do not ensure legally operative oversight findings can be recovered. This technical report defines an evidentiary-adequacy criterion for a bounded class of binary, event-specific determinations about relations such as boundary crossings, intervention possibilities, barrier integrity, and delegated authority validity. A runtime record answers only when it includes typing mapping recorded events to the legally operative category and the relation—provenance, authority, derivation, or temporal validity—upon which truth depends; necessity is proved by construction with minimality and bounded adequacy, applied to EU AI Act obligations.","arXiv :2607 .0094 1v 1 [ cs .CY] 1 Jul 2026  \nFrom Runtime Records to Legal Findings: An Evidentiary-Adequacy Criterion for Agentic AI Oversight  \nJeroen Janssen  \nApparens  \n[office@apparens. nl](office@apparens. nl)  \nTechnical report, July 2026  \nAbstract  \nAgentic AI systems generate runtime records, logs, traces, and audit artefacts, but the existence or integrity of such records does not by itself establish that legally operative oversight findings can be recovered from them. This technical report defines an evidentiaryadequacy criterion for a bounded class of determinations: binary findings of fact about specific events and their relations, such as whether protected data crossed a boundary, whether a human could intervene, whether an information barrier held, or whether delegated authority was valid at the moment of use. The criterion states that a runtime record can answer such a determination only if it carries both a typing that maps recorded events to the legally operative category and the relation, such as provenance, authority, derivation, or temporal validity, on which the determination’s truth depends. The claim is one of necessity, not sufficiency. It is shown by construction across the defined determination class, with minimality and bounded adequacy stated explicitly. The report instantiates the criterion against selected EU AI Act oversight obligations and explains why tamper-evident logs, generic process frameworks, and provenance structures alone cannot establish the relevant findings. It further relates the argument to requisite variety, the Good Regulator Theorem, and the trace-versus-hyperproperty boundary of runtime verification. Companion materials and the experiment protocol are archived on Zenodo.  \nKeywords: supervisory evidence; evidentiary adequacy; determination-answerability; runtime oversight; agentic AI; EU AI Act; provenance; hyperproperties; requisite variety; Good Regulator Theorem.  \n1 Introduction  \nAgentic AI systems select execution paths at run time. They use models, tools, memory, policies, and harnesses to compose sequences of actions that are not fully enumerated at design time. Enterprise governance has often responded by extending static instruments: point-in-time audits, conformity checklists, periodic review cycles, tamper-evident logs, and retrospective assurance reports. These instruments can be useful, but they do not answer a prior question: under what conditions can a record be read as evidence for a legally operative finding of fact?  \nThis report defines an evidentiary-adequacy criterion for runtime oversight. For a class of determinations that are findings of fact about specific events and relations, a runtime record can answer the determination only if the record carries two features: a typing that maps recorded events to the legally operative category, and a relation, typically provenance, derivation, authority, or temporal validity, on which the determination’s truth depends. The criterion is a necessity claim. It is not a claim that any such representation is sufficient for good oversight, compliance, truthfulness, or governance quality.  \nThe worked legal instance is EU AI Act oversight, especially the relation between logging and human oversight obligations. The paper’s legal claim is conditional: a record that is not  \nsemantically and relationally interpretable cannot serve as evidence that effective oversight was possible for determinations whose truth is not a function of the record. The report does not claim that the EU AI Act mandates a particular ontology or schema.  \n1.1 Contribution  \nThis report makes four contributions. First, it states a determination-answerability criterion over runtime evidence representations. Second, it demonstrates necessity by construction for four in-scope determinations and states minimality and bounded adequacy. Third, it positions the criterion against existing governance, provenance, logging, and runtime-monitoring approache","cbCaifYipW96hLGw","https://ap.wps.com/l/cbCaifYipW96hLGw","pdf",303543,1,12,"English","en",105,"# Introduction\n## Contribution\n## Scope and non-claims\n# The Evidentiary-Adequacy Criterion","[{\"question\":\"Why don’t runtime logs and audit artefacts alone establish legally operative oversight findings?\",\"answer\":\"Because the existence or integrity of records by itself does not guarantee that the legally operative determination’s truth can be recovered from the record. The report introduces a criterion specifying additional structural interpretability requirements.\"},{\"question\":\"What makes a runtime record answer a legally operative determination under the proposed criterion?\",\"answer\":\"A record answers only if it carries (1) typing that maps recorded events to the legally operative category and (2) the legally relevant relation such as provenance, authority, derivation, or temporal validity that the determination’s truth depends on.\"},{\"question\":\"Which kinds of determinations are in scope, and which are excluded?\",\"answer\":\"In scope are binary findings of fact about specific events and their relations (yes/no questions tied to particular occurrences). Out of scope are evaluative or statistical properties over behavior distributions, including fairness, robustness, non-discrimination, explainability, and transparency as stated.\"}]",1784181579,30,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"from-runtime-records-to-legal-findings-an-evidentiary-adequacy-criterion-for-agentic-ai-oversight","",{"@graph":35,"@context":84},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/from-runtime-records-to-legal-findings-an-evidentiary-adequacy-criterion-for-agentic-ai-oversight/82570/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"Why don’t runtime logs and audit artefacts alone establish legally operative oversight findings?","Question",{"text":74,"@type":75},"Because the existence or integrity of records by itself does not guarantee that the legally operative determination’s truth can be recovered from the record. The report introduces a criterion specifying additional structural interpretability requirements.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"What makes a runtime record answer a legally operative determination under the proposed criterion?",{"text":79,"@type":75},"A record answers only if it carries (1) typing that maps recorded events to the legally operative category and (2) the legally relevant relation such as provenance, authority, derivation, or temporal validity that the determination’s truth depends on.",{"name":81,"@type":72,"acceptedAnswer":82},"Which kinds of determinations are in scope, and which are excluded?",{"text":83,"@type":75},"In scope are binary findings of fact about specific events and their relations (yes/no questions tied to particular occurrences). Out of scope are evaluative or statistical properties over behavior distributions, including fairness, robustness, non-discrimination, explainability, and transparency as stated.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,121,126,129,133],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":28,"slug":120},"research-report",{"id":122,"doc_module":4,"doc_module_name":45,"category_name":123,"show_sort_weight":124,"slug":125},9,"Religion & Spirituality",20,"religion-spirituality",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":127,"show_sort_weight":124,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":45,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":45,"category_name":135,"show_sort_weight":105,"slug":136},19,"General","general"]