Skip to content

EU AI Act mapping

Traceability shortcuts: Taxonomy → Minimum Evidence → Validator → Human Oversight Protocol.

This page maps selected EU AI Act themes (documentation, record-keeping, risk management, human oversight, transparency) to AIMO evidence and artifacts. It is high-level only and does not constitute legal advice or guarantee conformity. Verify against the official legal text.

Reference: Regulation (EU) 2024/1689 (Artificial Intelligence Act). All article numbers below refer to that regulation.

Mapping table

Framework reference / topic AIMO evidence / where in AIMO Evidence Bundle / Minimum Evidence Artifacts & validation Notes
Art 4 – AI literacy Scope Summary, EV; review templates/ev/ Cross-cutting; evidence of organisational capability/training (high-level). Not legal advice. Verify against official text.
Art 9 – Risk management system Scope request, review, exception, renewal templates/ev/ High-risk AI systems (Title III). Not legal advice. Verify against official text.
Art 10 – Data and data governance Dictionary Dictionary, EV schemas/jsonschema/; schema_validate_dictionary Not legal advice. Verify against official text.
Art 11 – Technical documentation (high-risk) EV Template, Evidence Bundle EV, Dictionary, Summary; request, review schemas/jsonschema/, templates/ev/; Annex IV: see Examples > EU Annex IV sample (examples/evidence_bundle_v01_annex_iv_sample/); profile: coverage_map/profiles/eu_ai_act_annex_iv.json. Sample bundle is normative-compliant (signatures/, hashes/, payload with Annex IV–oriented technical documentation). See Examples for details (further sample content in a future release). High-level only; not legal advice. Verify against official text.
Art 12 – Record-keeping Evidence Bundle, Minimum Evidence EV, change_log, request, review examples/evidence_bundle_minimal/; schema_validate_ev Not legal advice. Verify against official text.
Art 13 – Transparency and provision of information to deployers/users Scope Summary, EV; review templates/ev/ High-risk context. Not legal advice. Verify against official text.
Art 14 – Human oversight Minimum Evidence review, exception templates/ev/ev_template.md Not legal advice. Verify against official text.
Art 15 – Accuracy, robustness, cybersecurity Minimum Evidence EV (evidence codes / risk codes, high-level) templates/ev/ High-level mapping only. Not legal advice. Verify against official text.
Art 17 – Quality management system Scope Summary, review (organisation process) templates/ev/ Distinct from Art 9 (risk management system). Not legal advice. Verify against official text.
Transparency obligations (use-case dependent) Scope, Minimum Evidence Summary, EV; review templates/ev/ Applicable provisions depend on use case (e.g. limited-risk, deployer duties). Not legal advice. Verify against official text.
GPAI models obligations EV Template, Evidence Bundle EV Template, Evidence Bundle (evidence-structuring framework) schemas/jsonschema/; schema_validate_ev AIMO provides a framework for organising evidence; actual obligations are defined by the regulation. Not legal advice. Verify against official text.
Recitals – Accountability Evidence Bundle EV, request, review, change_log examples/evidence_bundle_minimal/; schema_validate_ev Not legal advice. Verify against official text.

Effective dates / applicability (high-level)

The following aligns with EU official timeline (AI Act Service Desk / Commission). It is not legal advice and does not guarantee accuracy. Always confirm with the official legal text and competent authorities.

Phase Date What applies (high-level)
Entry into force August 2024 Regulation in force; most substantive obligations not yet applicable.
General provisions & prohibitions 02 Feb 2025 Prohibited practices (unacceptable risk); certain AI literacy–related provisions.
GPAI rules + governance 02 Aug 2025 Rules on notified bodies, GPAI, governance, confidentiality, penalties; codes of practice.
Majority rules + Annex III high-risk + Art 50 transparency 02 Aug 2026 Full applicability for high-risk AI systems (Annex III), Article 50 transparency obligations.
High-risk embedded in regulated products 02 Aug 2027 High-risk AI systems embedded in products subject to EU product legislation.

Harmonised standards and presumption of conformity (Article 40)

When harmonised standards are published in the EU Official Journal under the AI Act, compliance with them can provide presumption of conformity with the corresponding requirements. The exact list and dates depend on standardisation work and OJ publication. AIMO mappings are informative and do not confer presumption of conformity. For current status, see the European Commission AI Act standardisation page and References below.

2026 AI Office guidelines (implementation detail)

The European Commission has indicated that the AI Office will prepare practical guidelines during 2026, including on:

  • High-risk classification
  • Article 50 (transparency) implementation
  • Incident reporting
  • QMS-related elements

These guidelines are update triggers for AIMO profiles and coverage mappings: as they are published, adopters should align evidence and mappings with the latest official guidance. AIMO does not interpret or guarantee compliance with these guidelines.

!!! warning "Not legal advice" This page is for explanatory use only. You must verify applicability and dates against the official regulation and any implementing or amending acts. AIMO does not provide legal advice or guarantee compliance.

!!! note "Legal note / Informative mapping" This page is informative only. Legal interpretation should be based on the official regulation (EUR-Lex) and European Commission publications. AIMO does not provide legal advice or guarantee compliance.

References

Primary sources

Other

  • European Commission / AI Office — guidelines and timeline (see AI Act Service Desk and Commission news for current URLs)
  • EPRS — EU AI Act implementation — Parliament briefing (informative)