{"$schema":"https://policywindow.org/critique/api/schema","generated_by":"agi-social-scientist","url":"https://policywindow.org/critique/r/ai-governance-frameworks","attestation":{"ok":true,"checks":[{"id":"coverage","label":"Coverage matrix re-derived from codings matches the published aggregate","pass":true,"detail":"16 codings → 5×6 matrix"},{"id":"gaps","label":"Re-derived empty cells are exactly the published gaps (identity, not just count)","pass":true,"detail":"21 re-derived; same cells; 21 published"},{"id":"included","label":"Unique included papers match the published inclusion count","pass":true,"detail":"16 re-derived vs 16 published"},{"id":"excluded","label":"Excluded records match the published exclusion count","pass":true,"detail":"13 re-derived vs 13 published"},{"id":"spans","label":"Every inclusion carries a verbatim rationale span (AGISS P1: ≥20 chars)","pass":true,"detail":"shortest span 52 chars (min 20)"}],"derived":{"coverage":{"conceptual_governance_model":{"conceptual_normative":4,"legal_doctrinal":0,"empirical_qualitative":0,"empirical_quantitative":0,"technical_review_survey":0,"review_synthesis":2},"ethics_principles_framework":{"conceptual_normative":1,"legal_doctrinal":0,"empirical_qualitative":0,"empirical_quantitative":0,"technical_review_survey":0,"review_synthesis":0},"statutory_regulatory_framework":{"conceptual_normative":0,"legal_doctrinal":1,"empirical_qualitative":0,"empirical_quantitative":0,"technical_review_survey":0,"review_synthesis":0},"national_or_institutional_framework":{"conceptual_normative":1,"legal_doctrinal":0,"empirical_qualitative":0,"empirical_quantitative":0,"technical_review_survey":0,"review_synthesis":2},"sector_application_framework":{"conceptual_normative":2,"legal_doctrinal":0,"empirical_qualitative":0,"empirical_quantitative":1,"technical_review_survey":0,"review_synthesis":2}},"totalCodings":16,"uniqueIncluded":16,"emptyCells":21,"excluded":13,"minSpanLength":52},"reportHash":"befadbb8963c3084","corpusHash":"b6f941a82b8594df"},"id":"ai-governance-frameworks","reviewId":"CR-REV-001","agissStem":"framework-ai","question":"What does the scholarly literature report on AI governance frameworks?","title":"Evidence gap map: AI governance frameworks in the scholarly literature","reviewKind":"evidence_gap_map","method":"scoping_review_v1","headline":"Across a 5x6 framework matrix, 21 of 30 cells have no papers in this corpus of 16 included records; the populated cells concentrate in aspect 'conceptual_governance_model' (6 papers) and evidence type 'conceptual_normative' (8 papers).","verdict":"evidence_gap_map conducted over 29 records (16 included, 13 excluded with reasons): a coverage map with 21 named empty cells. Counts only — no importance adjudication (Sacred Rule 9); the report re-derives offline from corpus + codings.","coderModelId":"in-session:claude-agent","codingFrame":{"aspects":["conceptual_governance_model","ethics_principles_framework","statutory_regulatory_framework","national_or_institutional_framework","sector_application_framework"],"evidenceTypes":["conceptual_normative","legal_doctrinal","empirical_qualitative","empirical_quantitative","technical_review_survey","review_synthesis"]},"coverage":{"conceptual_governance_model":{"conceptual_normative":4,"empirical_qualitative":0,"empirical_quantitative":0,"legal_doctrinal":0,"review_synthesis":2,"technical_review_survey":0},"ethics_principles_framework":{"conceptual_normative":1,"empirical_qualitative":0,"empirical_quantitative":0,"legal_doctrinal":0,"review_synthesis":0,"technical_review_survey":0},"national_or_institutional_framework":{"conceptual_normative":1,"empirical_qualitative":0,"empirical_quantitative":0,"legal_doctrinal":0,"review_synthesis":2,"technical_review_survey":0},"sector_application_framework":{"conceptual_normative":2,"empirical_qualitative":0,"empirical_quantitative":1,"legal_doctrinal":0,"review_synthesis":2,"technical_review_survey":0},"statutory_regulatory_framework":{"conceptual_normative":0,"empirical_qualitative":0,"empirical_quantitative":0,"legal_doctrinal":1,"review_synthesis":0,"technical_review_survey":0}},"gaps":["conceptual_governance_model x legal_doctrinal: 0 papers","conceptual_governance_model x empirical_qualitative: 0 papers","conceptual_governance_model x empirical_quantitative: 0 papers","conceptual_governance_model x technical_review_survey: 0 papers","ethics_principles_framework x legal_doctrinal: 0 papers","ethics_principles_framework x empirical_qualitative: 0 papers","ethics_principles_framework x empirical_quantitative: 0 papers","ethics_principles_framework x technical_review_survey: 0 papers","ethics_principles_framework x review_synthesis: 0 papers","statutory_regulatory_framework x conceptual_normative: 0 papers","statutory_regulatory_framework x empirical_qualitative: 0 papers","statutory_regulatory_framework x empirical_quantitative: 0 papers","statutory_regulatory_framework x technical_review_survey: 0 papers","statutory_regulatory_framework x review_synthesis: 0 papers","national_or_institutional_framework x legal_doctrinal: 0 papers","national_or_institutional_framework x empirical_qualitative: 0 papers","national_or_institutional_framework x empirical_quantitative: 0 papers","national_or_institutional_framework x technical_review_survey: 0 papers","sector_application_framework x legal_doctrinal: 0 papers","sector_application_framework x empirical_qualitative: 0 papers","sector_application_framework x technical_review_survey: 0 papers"],"disclosedFragilities":["scoping retrieval with fixed queries, not a systematic search (coverage is query-bounded)","coding from abstracts only — full texts were not consulted; cells count papers, not extracted effect estimates","single-annotator coding (in-session:claude-agent); no second coder, no kappa","corpus bounded to 29 retrieved records; counts are corpus-relative, not field-level claims"],"nCorpus":29,"nIncluded":16,"nExcluded":13,"reportHash":"befadbb8963c3084","corpusHash":"b6f941a82b8594df","selection":{"method":"review_selector_v1","selectedKind":"evidence_gap_map","features":{"corpus_kind":"mixed","full_texts_available":false,"has_quantitative_effects":false,"n_studies":29,"n_with_abstracts":24,"outcomes_comparable":false},"selectionHash":"f41ed8259d2bd65e","verdict":"Review-kind selection for 'What does the scholarly literature report on AI governance frameworks?': SELECTED evidence_gap_map; 0 kind(s) rejected with their failed requirements recorded. A methodological screen (Sacred Rule 9): the selection is disclosed on the article and the selected kind's own discipline still applies at conduct time."},"includedCodings":[{"openalexId":"W2768432182","title":"A Layered Model for AI Governance","venue":"IEEE Internet Computing","year":2017,"doi":"10.1109/mic.2017.4180835","aspect":"conceptual_governance_model","evidenceType":"conceptual_normative","rationaleSpan":"this article proposes a conceptual framework for thinking about governance for AI"},{"openalexId":"W2896245505","title":"Governing artificial intelligence: ethical, legal and technical opportunities and challenges","venue":"Philosophical Transactions of the Royal Society A Mathematical Physical and Engineering Sciences","year":2018,"doi":"10.1098/rsta.2018.0080","aspect":"conceptual_governance_model","evidenceType":"review_synthesis","rationaleSpan":"This paper is the introduction to the special issue entitled: 'Governing artificial intelligence: ethical, legal and technical opportunities and challenges'"},{"openalexId":"W2897178686","title":"Ethical governance is essential to building trust in robotics and artificial intelligence systems","venue":"Philosophical Transactions of the Royal Society A Mathematical Physical and Engineering Sciences","year":2018,"doi":"10.1098/rsta.2018.0085","aspect":"conceptual_governance_model","evidenceType":"conceptual_normative","rationaleSpan":"as a framework to guide ethical governance in robotics and AI"},{"openalexId":"W2902634493","title":"AI4People—An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations","venue":"Minds and Machines","year":2018,"doi":"10.1007/s11023-018-9482-5","aspect":"ethics_principles_framework","evidenceType":"conceptual_normative","rationaleSpan":"present a synthesis of five ethical principles that should undergird its development and adoption"},{"openalexId":"W3017205423","title":"The Dark Sides of Artificial Intelligence: An Integrated AI Governance Framework for Public Administration","venue":"International Journal of Public Administration","year":2020,"doi":"10.1080/01900692.2020.1749851","aspect":"national_or_institutional_framework","evidenceType":"review_synthesis","rationaleSpan":"The literature provides few answers to the question of how government and public administration should respond"},{"openalexId":"W3094793347","title":"A strategic framework for artificial intelligence in marketing","venue":"Journal of the Academy of Marketing Science","year":2020,"doi":"10.1007/s11747-020-00749-9","aspect":"sector_application_framework","evidenceType":"conceptual_normative","rationaleSpan":"The authors develop a three-stage framework for strategic marketing planning"},{"openalexId":"W4214629857","title":"Defining organizational AI governance","venue":"AI and Ethics","year":2022,"doi":"10.1007/s43681-022-00143-x","aspect":"conceptual_governance_model","evidenceType":"conceptual_normative","rationaleSpan":"A concise AI governance definition would allow researchers and practitioners to identify the constituent parts of the complex problem of translating AI ethics into practice"},{"openalexId":"W4304943299","title":"Ethical principles for artificial intelligence in education","venue":"Education and Information Technologies","year":2022,"doi":"10.1007/s10639-022-11316-w","aspect":"sector_application_framework","evidenceType":"conceptual_normative","rationaleSpan":"the adoption of AIED has led to increasing ethical risks and concerns regarding several aspects"},{"openalexId":"W4315880904","title":"Ethics and governance of trustworthy medical artificial intelligence","venue":"BMC Medical Informatics and Decision Making","year":2023,"doi":"10.1186/s12911-023-02103-9","aspect":"sector_application_framework","evidenceType":"review_synthesis","rationaleSpan":"We adopted a multidisciplinary approach and summarized five subjects that influence the trustworthiness of medical AI"},{"openalexId":"W4317748910","title":"Artificial Intelligence Risk Management Framework (AI RMF 1.0)","venue":"","year":2023,"doi":"10.6028/nist.ai.100-1","aspect":"national_or_institutional_framework","evidenceType":"conceptual_normative","rationaleSpan":"the goal of the AI RMF is to offer a resource to the organizations designing, developing, deploying, or using AI systems to help manage the many risks of AI"},{"openalexId":"W4320713255","title":"On the Development of AI Governance Frameworks","venue":"IEEE Internet Computing","year":2023,"doi":"10.1109/mic.2022.3186030","aspect":"conceptual_governance_model","evidenceType":"review_synthesis","rationaleSpan":"This article explores new frameworks, institutional arrangements, and different types of regulation to devise and implement governance of artificial intelligence technologies, services, and devices"},{"openalexId":"W4383312437","title":"A comprehensive AI policy education framework for university teaching and learning","venue":"International Journal of Educational Technology in Higher Education","year":2023,"doi":"10.1186/s41239-023-00408-3","aspect":"sector_application_framework","evidenceType":"empirical_quantitative","rationaleSpan":"Data was collected from 457 students and 180 teachers and staff across various disciplines in Hong Kong universities, using both quantitative and qualitative research methods"},{"openalexId":"W4387002104","title":"Black box no more: a scoping review of AI governance frameworks to guide procurement and adoption of AI in medical imaging and radiotherapy in the UK","venue":"British Journal of Radiology","year":2023,"doi":"10.1259/bjr.20221157","aspect":"sector_application_framework","evidenceType":"review_synthesis","rationaleSpan":"AI governance frameworks are still under development"},{"openalexId":"W4389546926","title":"Trustworthy Artificial Intelligence: Design of AI Governance Framework","venue":"Strategic Analysis","year":2023,"doi":"10.1080/09700161.2023.2288994","aspect":"conceptual_governance_model","evidenceType":"conceptual_normative","rationaleSpan":"The focal point of the article is the Adaptive-Hybrid AI Governance framework based on technical, ethical, and societal regulatory mechanisms"},{"openalexId":"W4400849205","title":"The EU's AI act: A framework for collaborative governance","venue":"Internet of Things","year":2024,"doi":"10.1016/j.iot.2024.101291","aspect":"statutory_regulatory_framework","evidenceType":"legal_doctrinal","rationaleSpan":"It establishes a comprehensive legal framework with a risk-based approach"},{"openalexId":"W4406488361","title":"Governing intelligence: Singapore’s evolving AI governance framework","venue":"Cambridge Forum on AI Law and Governance","year":2025,"doi":"10.1017/cfl.2024.12","aspect":"national_or_institutional_framework","evidenceType":"review_synthesis","rationaleSpan":"This paper provides an outline analysis of the evolving governance framework for artificial intelligence (AI) in Singapore"}],"excludedPapers":[{"openalexId":"W2969625533","title":"Artificial Intelligence (AI): Multidisciplinary perspectives on emerging challenges, opportunities, and agenda for research, practice and policy","venue":"International Journal of Information Management","year":2019,"doi":"10.1016/j.ijinfomgt.2019.08.002","exclusionReason":"off-topic: broad AI overview; no governance framework proposed or analyzed"},{"openalexId":"W2981731882","title":"Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI","venue":"Information Fusion","year":2019,"doi":"10.1016/j.inffus.2019.12.012","exclusionReason":"no abstract available for coding (abstracts-only protocol)"},{"openalexId":"W2981863007","title":"Systematic review of research on artificial intelligence applications in higher education – where are the educators?","venue":"International Journal of Educational Technology in Higher Education","year":2019,"doi":"10.1186/s41239-019-0171-0","exclusionReason":"off-topic: review of AI applications in higher education"},{"openalexId":"W3000603264","title":"The role of artificial intelligence in achieving the Sustainable Development Goals","venue":"Nature Communications","year":2020,"doi":"10.1038/s41467-019-14108-y","exclusionReason":"off-topic: AI effects on SDGs; no governance framework proposed or analyzed"},{"openalexId":"W3023997891","title":"Artificial intelligence and the future of global health","venue":"The Lancet","year":2020,"doi":"10.1016/s0140-6736(20)30226-9","exclusionReason":"off-topic: AI prospects in global health; no governance framework"},{"openalexId":"W3035286874","title":"Ethical and legal challenges of artificial intelligence-driven healthcare","venue":"Elsevier eBooks","year":2020,"doi":"10.1016/b978-0-12-818438-7.00012-5","exclusionReason":"no abstract available for coding (abstracts-only protocol)"},{"openalexId":"W3036911563","title":"Data governance: Organizing data for trustworthy Artificial Intelligence","venue":"Government Information Quarterly","year":2020,"doi":"10.1016/j.giq.2020.101493","exclusionReason":"no abstract available for coding (abstracts-only protocol)"},{"openalexId":"W3084021717","title":"Vision, challenges, roles and research issues of Artificial Intelligence in Education","venue":"Computers and Education Artificial Intelligence","year":2020,"doi":"10.1016/j.caeai.2020.100001","exclusionReason":"off-topic: vision paper on AI in education"},{"openalexId":"W3147517805","title":"Human- versus Artificial Intelligence","venue":"Frontiers in Artificial Intelligence","year":2021,"doi":"10.3389/frai.2021.622364","exclusionReason":"off-topic: human-vs-AI intelligence comparison"},{"openalexId":"W3156614709","title":"A Review of Artificial Intelligence (AI) in Education from 2010 to 2020","venue":"Complexity","year":2021,"doi":"10.1155/2021/8812542","exclusionReason":"off-topic: content analysis of AI-in-education research"},{"openalexId":"W3204486714","title":"Ethical framework for Artificial Intelligence and Digital technologies","venue":"International Journal of Information Management","year":2021,"doi":"10.1016/j.ijinfomgt.2021.102433","exclusionReason":"no abstract available for coding (abstracts-only protocol)"},{"openalexId":"W4293229829","title":"Unlocking the value of artificial intelligence in human resource management through AI capability framework","venue":"Human Resource Management Review","year":2022,"doi":"10.1016/j.hrmr.2022.100899","exclusionReason":"no abstract available for coding (abstracts-only protocol)"},{"openalexId":"W4379470483","title":"A Review of the Role of Artificial Intelligence in Healthcare","venue":"Journal of Personalized Medicine","year":2023,"doi":"10.3390/jpm13060951","exclusionReason":"off-topic: role-of-AI-in-healthcare review; no governance framework"}],"narrativeMarkdown":"# Evidence gap map: AI governance frameworks in the scholarly literature\n\n## In plain terms\n\n\"AI governance framework\" is one of the most-used phrases in the field — but what kinds of frameworks does the scholarship actually contain, and on what evidence do they stand? The engine retrieved a corpus on the question, coded every abstract into a framework matrix, and counted. The picture: a literature rich in conceptual models and principle sets, thin on statute, and almost empty of empirical testing — 21 of 30 matrix cells contain no papers at all.\n\n**Finding (screened coverage map):** Across a 5x6 framework matrix, 21 of 30 cells have no papers in this corpus of 16 included records; the populated cells concentrate in aspect 'conceptual_governance_model' (6 papers) and evidence type 'conceptual_normative' (8 papers).\n\n## Background\n\nPolicy Window's research engine selected the review kind BEFORE this article was drafted: a review-kind selector matched the question (\"What does the scholarly literature report on AI governance frameworks?\") and the measured corpus features (29 records, 24 with abstracts, no comparable quantitative effects, abstracts only) to an evidence gap map (selection `f41ed8259d2bd65e`). The topic is the engine's own study queue's top unserviced entry. The article's status is a screened candidate routed to human review, not an adjudicated truth (Sacred Rule 9).\n\n## Method\n\nEvery record was coded exactly once against a declared frame — five framework aspects (conceptual models, ethics-principles sets, statutory frameworks, national/institutional frameworks, sector application frameworks) by six evidence types — with a verbatim rationale span per inclusion and a stated reason per exclusion (13 records excluded: off-topic or no abstract). The coverage table is arithmetic over those codings. Review report `befadbb8963c3084`; corpus `b6f941a82b8594df`.\n\n## What the map shows\n\nThe populated region: conceptual governance models (layered models, roadmaps, definitional work, hybrid designs), national and institutional frameworks (the NIST AI Risk Management Framework, Singapore's evolving framework, public-administration designs), one statutory entry (the EU AI Act analyzed doctrinally), and sector application frameworks in health, education, and marketing. The named empty cells include nearly every aspect crossed with empirical evidence: one included paper reports a quantitative survey, none reports qualitative empirical study of a framework in use, and no framework aspect carries more than a single empirical paper in this corpus. Frameworks are proposed far more often than they are tested.\n\n## Limitations (disclosed by the engine)\n\n- scoping retrieval with fixed queries, not a systematic search (coverage is query-bounded)\n- coding from abstracts only — full texts were not consulted; cells count papers, not extracted effect estimates\n- single-annotator coding (in-session:claude-agent); no second coder, no kappa\n- corpus bounded to 29 retrieved records; counts are corpus-relative, not field-level claims\n\n## Verify\n\nEvery count above re-derives offline from the committed corpus and codings — no model, no network, no trust in the institute required: `PYTHONPATH=src python scripts/verify_review_framework_ai.py  # exit 0 = re-derives offline`\n","verifyCommand":"PYTHONPATH=src python scripts/verify_review_framework_ai.py  # exit 0 = re-derives offline","aiAgiCategories":["AI_governance","knowledge_production"],"publicationDate":"2026-06-14","published":true}