Every AI Policy Ever · open in the interactive register

Published (non-binding, current)Non-binding · Guidelines / framework

Methodology for the Risk and Impact Assessment of Artificial Intelligence Systems from the Point of View of Human Rights, Democracy and the Rule of Law (HUDERIA Methodology) and HUDERIA Model

Record COE-02

Issued by
Council of Europe Committee on Artificial Intelligence (CAI)
Jurisdiction
Council of Europe
Level · region
International · Europe
Published / adopted
2024-11-28
Last updated
2026-02-25
Language
English, French

Summary

Non-binding Council of Europe guidance for assessing the risks and impacts of AI systems on human rights, democracy and the rule of law across their lifecycle, for public and private actors. The Methodology sets out a context-based risk analysis (COBRA) with scoping, risk factors, impact mapping and triage; a stakeholder engagement process; a risk and impact assessment; and a mitigation plan, with iterative review. The Model's COBRA Resources list risk factors in an AI system's application, design and development, and deployment contexts, and illustrative areas of concern and sectors. Parties to the Framework Convention may use or adapt it; it is not a means of implementing the Convention.

Safety topics

risk-based classificationhuman oversightaccountability & liability

Open question

date_last_updated is unsupported and the checker gave no evidence-backed fix Issue #115

The record shows its last checked values until a maintainer decides; the decision is recorded in its history.

Status notes

The CAI adopted the HUDERIA Methodology at its 12th plenary meeting on 26-28 November 2024, and the HUDERIA Model's Context Based Risk Analysis (COBRA) Resources (CAI(2025)6rev2, dated 5 November 2025) at its 14th plenary meeting on 3-5 November 2025; the Commission's note forwarding the Model to the Council gives 6 November 2025 as the date of that adoption. The Methodology was adopted first, and the detailed supporting material (questionnaires, problematic uses, methodological guidance) was postponed to 2025 as the Model. Council of Europe Publishing issued both parts together in 2026 as 'HUDERIA - Methodology and Model', in English and French. The CAI's work on HUDERIA was to continue in its successor committee, the Steering Committee on New and Emerging Technologies (CDNET).

Sources

Open the official source https://www.coe.int/en/web/artificial-intelligence/huderia-risk-and-impact-assessment-of-ai-systems

Secondary source: https://book.coe.int/en/artificial-intelligence/12519-huderia-methodology-and-model.html

Local copy

Partial (secondary source only). The official source could not be downloaded (HTTP 403 Forbidden); only the secondary source was saved. The browser retry also failed (HTTP 403 (in browser)). Last attempt 2026-10-17 18:17:55 UTC

Secondary source (third-party copy, not published) archive/COE-02/2026-10-17/secondary.html
Retrieved 2026-10-17 18:18:05 UTC
SHA-256 54d6beedb91e0d33ecf0cf5bf7408ef303a5952686b524c09f07cf338555fc92

Verification

A
The instrument's own official text, or an official register, was stored and confirms the title, date and status. checked 2026-10-17 · official-url-fetched
Stored and read in this run: the Council of Europe Publishing page 'HUDERIA - Methodology and Model (2026)' (book.coe.int), which confirms the title of the combined publication, its development by the CAI, its description as a non-legally binding instrument and the table of contents behind the summary; Council of the EU documents 16566/25 and 16537/25 (data.consilium.europa.eu), which give the CAI's adoption of the Methodology at its 12th plenary (26-28 November 2024) and of the COBRA Resources at its 14th plenary (3-5 November 2025), the non-binding character and the document reference CAI(2025)6rev2; and the addendum 16537/25 ADD 1, a scan read through its OCR text layer, which reproduces the HUDERIA Model: COBRA Resources (CAI(2025)6rev2) with its statement that HUDERIA is non-legally binding guidance without legal effect. The record's own sources, the coe.int HUDERIA page and the rm.coe.int PDF of the combined text, are not stored: the sites answered HTTP 403 to the archiver, in the plain download and in the browser retry. The Committee of Ministers' approvals are therefore not confirmed.
Second check: Independent check by Claude Code (claude-sonnet-5-5), 2026-10-17: 6 field(s) supported.

History

  1. 2026-10-18related: [{"slug": "coe-framework-convention-ai", "id": "COE-01"}] → [{"slug": "coe-framework-convention-ai", "id": "COE-01"}, {"id": "SI-02", "slug" — two-way link added: SI-02 link(s) to COE-022026-10-18-fix-related-links · Claude Code (claude-opus-5-5)
  2. 2026-10-17notes: Methodology adopted by the CAI on 28 Nov 2024 and approved by the Committee of M → The CAI adopted the HUDERIA Methodology at its 12th plenary meeting on 26-28 Nov — rewritten from the stored texts; the plenary dates, CAI(2025)6rev2 and the 5 vs 6 November 2025 discrepancy added; the Committee of Ministers' approvals and the 2026 pilots moved to open_questions2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)
  3. 2026-10-17safety_topics: ["risk-based classification", "accountability & liability", "public-sector use"] → ["risk-based classification", "human oversight", "accountability & liability"] — 'public-sector use' dropped (the guidance is for public and private actors alike); 'human oversight' added (enhanced human oversight as a compensating control in COBRA Resources A)2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)
  4. 2026-10-17summary: Non-binding guidance for assessing risks and impacts of AI systems on human righ → Non-binding Council of Europe guidance for assessing the risks and impacts of AI — rewritten from the publication's table of contents and the Model's preface; version 1's 'mitigation plan including remedies' and 'It supports implementation of the CoE Framework Convention' are not in the stored text, and the Model says HUDERIA 'is not a means to implement the Framework Convention'2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)
  5. 2026-10-17secondary_url: https://rm.coe.int/huderia-methodology-and-model-adopted-provisional-version-202 → https://book.coe.int/en/artificial-intelligence/12519-huderia-methodology-and-mo — the rm.coe.int PDF is not stored; the Council of Europe Publishing page is, so the record's local copy can hold an official page of the issuer2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)
  6. 2026-10-17issuing_body: Council of Europe Committee on Artificial Intelligence (CAI); approved by the Co → Council of Europe Committee on Artificial Intelligence (CAI) — 'approved by the Committee of Ministers' removed: no stored file records an approval by the Committee of Ministers; the CAI as author is supported2026-10-17-recheck-europe-1 · Claude Code (claude-opus-5-5)

Related records

CoE AI Convention Council of Europe · Adopted, not yet in force

Slovenia National AI Strategy 2030 (NsUI 2030) Slovenia · Published (non-binding, current)

Citation

Council of Europe Committee on Artificial Intelligence (CAI). Methodology for the Risk and Impact Assessment of Artificial Intelligence Systems from the Point of View of Human Rights, Democracy and the Rule of Law (HUDERIA Methodology) and HUDERIA Model. 2024-11-28. https://www.coe.int/en/web/artificial-intelligence/huderia-risk-and-impact-assessment-of-ai-systems. Status: Published (non-binding, current); verified 2026-10-17 (level A). Local copy archive/COE-02/2026-10-17/secondary.html retrieved 2026-10-17 18:18:05 UTC, SHA-256 54d6beedb91e0d33ecf0cf5bf7408ef303a5952686b524c09f07cf338555fc92. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record COE-02.

Improve this record

Know of an amendment, a status change or an error? Report news or an error on GitHub; the form opens with this record filled in.