EU AI Act & ISO/IEC
42001 readiness.
Plain-English classification for leaders and practical evidence for engineers. Answer three questions for an indicative first view of how the EU AI Act may apply to a system*.
Where does your AI system fall under the EU AI Act?
Three quick questions for an indicative view*.
What to expect
- Three questions or fewer, about 30 seconds.
- An indicative first view of where your system may sit.
- Nothing is stored or sent unless you choose to send your result.
EU AI Act timeline
Based on the European Commission AI Act Service Desk timeline, which reflects the Digital Omnibus on AI.
Prohibitions, GPAI and penalties
Prohibitions and AI literacy measures apply from 2 February 2025. Rules for general-purpose AI models, and penalty rules, apply from 2 August 2025.
Transparency and enforcement
Article 50 transparency duties apply and enforcement starts on 2 August 2026. New prohibitions apply, and the Article 50(2) marking transition for existing generative systems ends, on 2 December 2026.
Annex III high-risk
Obligations for standalone high-risk systems listed in Annex III apply from 2 December 2027.
Annex I product-embedded
Obligations for high-risk AI that is a safety component of products under EU product law (Annex I) apply from 2 August 2028.
Last reviewed 17 November 2026. Dates reflect the AI Act as amended by the Digital Omnibus on AI. Confirm your position with counsel.
One set of evidence, mapped to four frameworks.
Map each technical control once and reuse the evidence across the EU AI Act, ISO/IEC 42001, ISO/IEC 27001 and SOC 2 criteria.
| Control domain | EU AI Act (Regulation (EU) 2024/1689) | ISO/IEC 42001 (AIMS) | ISO/IEC 27001, SOC 2 and GDPR | What we set up |
|---|---|---|---|---|
| Risk management | Article 9 (high-risk systems) | Clause 6.1 (risk assessment and treatment) and the impact-assessment controls in Annex A.5 | ISO/IEC 27001 clause 6.1.2 / SOC 2 CC3.2 | Living AI risk register and risk checks in CI/CD |
| Data governance | Article 10 (high-risk systems) | Data-for-AI controls, Annex A.7 | GDPR Article 5 (accuracy and data minimisation) | Lineage, data quality tests and bias checks in the data pipeline |
| Technical documentation | Article 11 and Annex IV (high-risk systems) | Technical documentation within the life-cycle controls, Annex A.6 | SOC 2 CC8.1 (change management) | Annex IV documentation drafts, model cards and release records, reviewed by a person |
| Logging and traceability | Articles 12 and 19 (providers); Article 26(6) (deployers): at least six months | Event logging within the life-cycle controls, Annex A.6 | SOC 2 CC7.2; NIS2 where applicable | Event logs kept for at least six months where the Act applies, with access controls |
| Transparency and provenance | Article 50 (disclosure and marking) | Information for interested parties and users, Annex A.8 | GDPR Articles 13–14 | Disclosure text in interfaces; machine-readable marking of generated content |
| Human oversight and FRIA | Article 14 (oversight); Article 27 (FRIA, certain deployers) | Use of AI systems, Annex A.9 | GDPR Article 22 and Article 35 (DPIA) | Approval and override controls; FRIA and DPIA templates where required |
Article 99 penalty caps.
Maximum administrative fines under Article 99 of Regulation (EU) 2024/1689, from the official text. These are caps. National authorities set the amount within them, considering the factors in Article 99(7).
Up to €35 million or 7% of total worldwide annual turnover, whichever is higher, for prohibited AI practices (Article 99(3)). For SMEs, the lower of the two applies (Article 99(6)); for small mid-cap enterprises, the same rule applies to the €15 million and €7.5 million fines (Article 99(6a)).
Up to €15 million or 3% of worldwide annual turnover for breaches of provider and deployer obligations, including Article 50 transparency duties (Article 99(4)).
Up to €7.5 million or 1% for supplying incorrect, incomplete or misleading information to notified bodies or national authorities in reply to a request (Article 99(5)).
Penalty rules have applied since 2 August 2025. Article 101 fines for general-purpose AI model providers apply from 2 August 2026.
* Indicative self-assessment based on Regulation (EU) 2024/1689, as amended. Not legal advice. Check your own position with qualified counsel.
