Featured Publicaiton, INSIGHTS AND RESEARCH
Position papers, regulatory submissions, framework updates, and short reads from inside the Institute. Published when we have something to say.
Subscribe below to get it in your inbox.
FEATURED PUBLICATIONS
First PreLiminary Report of the UN INDEPENDENT SCIENTIFIC PANEL on AI
A critical reading of the first Preliminary Report of the UN Independent International Scientific Panel on AI, and what it means for the EW-AiRM™ framework
By Prof. Markus Krebsz | Human-Ai.Institute | July 2026
FEATURED PUBLICATIONS
FORTHCOMING BOOK
Wiley Finance, 2026. The full framework reference. Three layers, six strategic pillars, the MIT-based operational layer, the eight-category AI Black Swan resilience model, and the HAIPECR ethical overlay.
Read about the book →
UNECE 2024
The UNECE WP.6 voluntary instrument for AI regulation across 56 member states. Marky is the project lead and author. Sits as the regulator-facing complement to EW-AiRM™ at the enterprise level.
UNECE website ↗
OECD LISTING, 2023
The seven-dimension ethical overlay has been listed on the OECD AI Policy Observatory since April 2023. Mapped to the UNESCO 2021 Recommendation on the Ethics of AI.
HAIPECR overview →
UN SUBMISSION, APR 2026
Co-authored paper (Mamun, Faride, Hansel, Krebsz) submitted to the UN Global Dialogue on AI Governance. Applies the three-framework stack (UNECE CRA, EW-AiRM™, HAIPECR) to sovereign AI in lower-income contexts.
Request a copy →
WORKING NOTES
A rolling set of working notes on AI governance, regulatory developments, and the design choices behind EW-AiRM™.
A short note on why typography carries weight in framework design. The "i" stands for inclusivity, deliberately lowercase to signal an embedded prerequisite rather than a standalone pillar.
Coming soon
The EU AI Act tells you what is prohibited and high-risk. EW-AiRM™ tells you how to run governance over what you keep. This note walks through the practical interlock between the two.
Coming soon · See FAQ for the short version
The framework’s resilience layer was expanded from seven to eight Black Swan categories to cover the NIST FIPS 203/204/205 transition. Why post-quantum is now an AI risk problem.
Coming soon
From Chapter 2 of the book. A short essay on why AI governance has to be deployable under uncertainty, and what changes when you accept the regulatory picture will keep moving for the next decade.
Coming soon
The operational layer of EW-AiRM™ is built on the MIT AI Risk Repository (CC BY 4.0). Why that choice matters, what it costs, and what it gives back.
Coming soon
An assessment of EW-AiRM™ against the Accountability Horizon paper. Identifies one place where the framework leans on an assumption that does not survive multi-agent deployments. Includes a proposed correction.
Coming soon
A short note when we publish something substantive, plus a quarterly framework update. No marketing. No tracking pixels. Unsubscribe in one click. Subscribe via the form below, or email subscribe@human-ai.institute directly.
The framework overview is the best one-page entry point. The book, when it lands, is the long-form version.
The Human-Ai.Institute est un Think/Do Tank indépendant qui organise un dialogue multilatéral et multipartite sur la gouvernance de l'IA. L'Institut collabore avec l'ONU, l'UNESCO, l'OCDE, la Commission européenne et des gouvernements du monde entier.
Accueil de EW-AiRM™ — le cadre ouvert à trois couches pour la gestion des risques liés à l'IA à l'échelle de l'entreprise, fondé sur des normes sous licence publique et aligné sur l'UNESCO, la CEE-ONU, le NIST, l'ISO et l'IA Act de l'UE.
Des services de conseil commercial, de formation et de conseil sont fournis par De-Risking Solutions Ltd. et RiskAi.Ai
SOLUTIONS IA ET TECHNOLOGIQUES
Les solutions d'IA, d'outils d'agents et technologiques sont fournies par Human-Ai.Solutions.
The Human-Ai.Institute
est géré et maintenu par
De-Risking Solutions Ltd.
E-mail :
contact [at] human-ai.institute
Listé sur l'Observatoire des politiques de l'IA de l'OCDE · Membre fondateur du Réseau IA de l'Université des Nations Unies · Alignement avec l'UNESCO, le WP.6 de la CEE-ONU, le NIST et l'ISO 42001
EW-AiRM™ et HAiPECR™ sont des marques déposées de De-Risking Solutions Ltd., enregistrées en Angleterre et au Pays de Galles (Numéro d'entreprise 09900565). Tous droits réservés.
© 2026 The Human-Ai.Institute.
