The AI Act marks a major step in the regulation of artificial intelligence in Europe. But, as with the GDPR in its early days, the text alone is not enough to fully understand the obligations it creates. Its implementation relies on an evolving set of guidelines, interpretative documents, and resources published by the European Commission and the AI Office.
In this context, keeping track of official publications is a key issue for lawyers, DPOs, compliance teams, and AI stakeholders. These documents help clarify the regulation’s scope, detail concrete obligations, and guide organizations in achieving compliance.
This article provides a complete mapping of the official European Union documents related to the AI Act, distinguishing between publications already available and those still expected.
⚠️ As this body of material is still being built, this mapping is updated regularly to include new resources as they are published.
You can find all of these official documents, centralized and regularly updated, directly in the “Resources” section of the Dastra platform, to simplify your regulatory monitoring and your AI Act compliance efforts.
AI Act: official EU texts and links
| Topic | Resource | Authority | Official link |
|---|---|---|---|
| AI Act text | Text of Regulation (EU) 2024/1689 – Artificial Intelligence Act | European Parliament & Council | AI Act link |
| Digital Omnibus AI text | Proposal for a Digital Omnibus Regulation on AI | European Parliament & Council | Digital Omnibus AI link |
| Main AI Act portal | “European approach to AI” page (resources, news, implementation) | European Commission | Portal link |
| Support & guidance | AI Act Service Desk (FAQ, tools, practical resources) | European Commission / AI Office | AI Act Service Desk link |
| Governance & oversight | Official AI Office page (mission, initiatives, publications) | European Commission (AI Office) | European AI Office |
| Terminology | Terminology and taxonomy of artificial intelligence | EU-U.S. Trade and Technology Council | Download link |
Soft law: published official documents (mapping by article)
AI Act | Topic | Associated | Publication | Official link | Dastra link |
|---|---|---|---|---|---|
Art. 3 (in force) | Definition of an AI system | Commission guidelines: definition of an AI system | 29/07/2025 | ||
Art. 5 (in force) | Prohibited practices | Commission guidelines: prohibited AI practices | 2029/07/25 | ||
Art. 50 (entry into force 2 August 2026) | Transparency (generative AI systems, chatbots, deepfakes) | First draft code of practice on marking and labeling AI-generated content | 17/12/2025 (partial) | ||
Art. 50 (entry into force 2 August 2026) | Transparency (generative AI systems, chatbots, deepfakes) | Second draft code of practice on marking and labeling AI-generated content | 03/03/2026 (partial) | ||
Art. 51–56 (in force) | General-purpose AI models (GPAI) | Code of practice (AI Office) | 10/07/2025 | ||
Art. 51–56 (in force) | General-purpose AI models (GPAI) | Commission guidelines: scope and obligations of GPAI models | 19/11/2025 |
Soft law: published official templates
| Topic | Template | Authority | Official link |
|---|---|---|---|
| Reporting serious incidents (systemic-risk GPAI) | Reporting template for serious incidents involving systemic-risk GPAI models | European Commission | Template to download |
| Training data transparency (GPAI) | Public summary template for training data content | European Commission | Template to download |
Upcoming documents and regulatory roadmap
| AI Act Article | Topic | Expected EU document | Status |
|---|---|---|---|
| Art. 6 + Annex III | High-risk systems | Guidelines on high risk AI Systems | ⏳ In progress, expected May/June 2026 |
| Art. 50 | Transparency | Code of practice | ⏳ In progress, expected June 2026 |
| Art. 9 | Risk management system | Harmonized standards + Commission guidance | ⏳ In progress |
| Art. 10 | Data governance | Guidelines on data quality and governance | ⏳ Expected |
| Art. 11 + Annex IV | Technical documentation | Commission templates and guidance | ⏳ Expected |
| Art. 12 | Logging | Guidance on traceability and monitoring | ⏳ Expected |
| Art. 14 | Human oversight | Guidelines on human oversight | ⏳ Expected |
| Art. 15 | Robustness, security, and accuracy | Harmonized standards | ⏳ Expected |
| Art. 16–29 | Provider obligations | Implementing acts | ⏳ To come |
| Art. 30–39 | Conformity assessment | Implementing acts | ⏳ To come |
From AI Act requirements to action: get started with Dastra’s AIMS
Although some guidelines and clarifications are still expected from the European Commission and the authorities, the AI Act has already entered into force. Organizations must therefore begin structuring their AI governance without waiting for a fully stabilized framework.
In practice, many still face the complexity of the regulation, the diversity of use cases, and a lack of an overall view. The result: fragmented initiatives, insufficiently controlled risks, and compliance that is difficult to demonstrate.
That is precisely what Dastra’s AI Management System (AIMS) is designed to address.
Gain access to a structured methodology, concrete use cases, and operational tools to start your AI Act compliance journey today.