European Union - AI Act Implementation
AI Board Sixth Meeting - Implementation and Interoperability
European Union
RAI-EU-NA-ABSMIXX-2025This policy guides EU national authorities and the European Commission on consistent AI Act implementation and technical interoperability, issued by the European Artificial Intelligence Board in 2025. In force since 4 December 2025, it establishes 2026 subgroup workplans and directs interim guidance for market surveillance across Member States.
Summary
On 4 December 2025 the European Artificial Intelligence Board (AI Board) convened for its sixth meeting to review current priorities and concrete next steps for AI Act implementation across Member States. The meeting focused on implementation and interoperability challenges arising from the risk-based framework in the AI Act and on short-term measures to ensure consistent application while voluntary and mandatory technical standards are still being finalised. The European Commission presented and discussed the Digital Omnibus package and the Data Union Strategy (published 19 November 2025) as instruments designed to simplify and streamline AI, cybersecurity and data rules and to unlock higher-quality datasets for AI, thereby supporting interoperable AI deployments across the Union.
Key outcomes included the Board's endorsement of 2026 workplans for several subgroups, an agreement to continue the rotating Council Presidency chairing model (handover from Denmark to Cyprus on 1 January 2026), and a Commission commitment to prepare interim guidance and Q&As to bridge gaps where international or European standards are delayed. Members exchanged national implementation experiences, shared good practices on national coordination and market surveillance, and discussed the interplay between the AI Act and other sectoral regimes (notably medical devices). The AI Office provided an update on technical developments in health and life sciences AI. The Board also reaffirmed its role in supporting conformity assessment, market surveillance coordination and in advising on delegated and implementing acts under the AI Act.
The meeting emphasised interoperability as both a technical and regulatory priority: interoperability requirements affect conformity assessment, data access and quality (including the Data Union Strategy), model evaluation and cross-border enforcement. To address these, the Commission signalled targeted simplification measures in the Digital Omnibus proposal and the intention to issue bridging guidelines, together with support via subgroups and regulatory sandboxes at EU and national level. The meeting output is primarily procedural and advisory (consistent with the AI Board’s role created by the AI Act) but carries practical effect because it directs Commission preparatory work (guidance, delegated/implementing acts) and coordination among national competent authorities. Primary official sources for these outcomes are the European Commission’s AI Board pages and the specific meeting note published by the European AI Office on the Commission’s Digital Strategy site.
Full article
Read full text ↗Overview
The sixth meeting of the European Artificial Intelligence Board was held on 4 December 2025 to focus on implementation and interoperability issues arising from the AI Act and associated secondary instruments. The meeting reviewed the Commission’s recent policy package (the Digital Omnibus proposal and the Data Union Strategy, both published 19 November 2025) and discussed immediate steps to ensure that obligations under the AI Act can be applied consistently across Member States while European and international standards are still being developed. The Board exchanged national experiences on enforcement and market surveillance, endorsed subgroup workplans for 2026 and agreed to continue the rotating chair model (handover from Denmark to Cyprus on 1 January 2026). Official meeting information and the Commission’s summary are available via the European Commission’s AI Office pages, including the meeting note published on the Commission’s Digital Strategy site (Sixth AI Board meeting, 4 December 2025).
Definitions
For the purposes of the Board’s meeting and this policy entry: "AI Board" means the European Artificial Intelligence Board established under the Artificial Intelligence Act (Regulation (EU) 2024/1689); "AI Office" means the European AI Office in the Commission acting as Secretariat; "interoperability" refers to technical, organisational and legal measures that enable AI systems, datasets, conformity assessment and oversight tools to function across borders and regulatory regimes; "Digital Omnibus" is the Commission initiative to streamline and simplify digital rules including targeted adjustments to the AI Act’s implementation timelines and requirements; "Data Union Strategy" denotes Commission actions to improve data availability and standardisation for AI. These terms are used consistently with the AI Act and Commission explanatory material.
Governance and Institutional Framework
The AI Board operates as the governance and coordination mechanism foreseen by the AI Act to assist Member States, coordinate national competent authorities and advise the Commission on delegated and implementing acts. It brings together representatives of Member States, the European AI Office (which acts as Secretariat), and observers such as the European Data Protection Supervisor and EEA/EFTA participants. At the sixth meeting the Board reaffirmed its core governance responsibilities: coordinating national enforcement and market surveillance approaches, endorsing subgroup workplans, and providing advice on the Commission’s guidance and delegated instruments. The Commission indicated it will prepare interim guidance documents to bridge gaps where standards are delayed, while standardisation bodies finalize technical norms. The Board’s administrative and advisory functions are documented on the Commission’s AI Board page (AI Board) and in the meeting note (Sixth AI Board meeting), which also direct stakeholders to the Register of Commission Expert Groups for agendas and minutes.
Key Focus Areas
The sixth meeting centred on several interlinked focus areas: first, practical implementation of the AI Act across Member States, with attention to how national competent authorities will coordinate conformity assessment, registration and market surveillance for high-risk systems. Second, interoperability — both technical (standards, data schemas, APIs, evaluation and testing methods) and regulatory (alignment of sectoral rules such as medical devices with AI obligations). Third, bridging the gap between law and standardisation: the Commission flagged the Digital Omnibus proposal as a mechanism for targeted simplification to ensure timely and proportionate implementation. Fourth, data availability and quality — the Board discussed the Commission’s Data Union Strategy to foster access to high-quality datasets necessary for compliant model development and evaluation. Fifth, sectoral updates especially in health and life sciences where the AI Office presented technical developments and the sub-group on interplay with the Medical Devices Regulation/IVDR continues to produce guidance. The Board also considered the role of voluntary tools such as the General-Purpose AI (GPAI) Code of Practice previously assessed by the Commission and the Board as an adequate demonstration tool where applicable.
Implementation Framework
The Board discussed a multi-layered implementation framework combining: (1) EU-level delegated and implementing acts, to be prepared by the Commission and subject to Board consultation; (2) interim Commission guidance and Q&As targeted to bridge delays in standards development; (3) subgroup activity to prepare sectoral guidance (e.g., health, medical devices interplay); (4) national implementation roadmaps by Member States and coordinated market surveillance; and (5) practical instruments such as regulatory sandboxes and repositories for test results and conformity documentation. The Commission signalled targeted simplifications in the Digital Omnibus proposal aimed at reducing administrative burdens while preserving core protections. The Board’s endorsement of 2026 subgroup workplans provides a roadmap for developing practical interoperability specifications, model evaluation methodologies and cross-border enforcement arrangements.
Monitoring and Evaluation
Monitoring at EU level will rely on coordinated reporting from national competent authorities, the AI Office's central functions and the AI Board subgroups. The meeting underlined the need for shared indicators on conformity assessment throughput, market surveillance actions, incident reports, and interoperability test outcomes. The Board agreed to strengthen channels for sharing data and evidence supporting enforcement and to use subgroup outputs and sandboxes to pilot evaluation methodologies. The Commission will use these inputs to inform delegated and implementing acts and, where appropriate, further adjustments via the Digital Omnibus exercise.
Penalties, Liability, and Appeals
The sixth meeting did not change the penalty or liability architecture set out in the AI Act itself; rather, it focused on practical enforcement coordination to ensure consistent application of the AI Act’s liability and sanctioning provisions at national level and on cross-border cases. The Board discussed the role of national competent authorities in applying administrative fines, corrective measures and orders as defined in the AI Act, and the importance of interoperability of incident reporting, conformity documentation and investigative tools to enable enforcement. The Board also addressed the need for transparent appeal routes and for Member States to align procedural safeguards in national enforcement mechanisms, while respecting the statutory rights and remedies enshrined in the AI Act and other Union law.
Relationship to Other Instruments
The Board examined the interplay between the AI Act and sectoral regimes, including the Medical Devices Regulation (MDR/IVDR), sectoral data rules and cybersecurity law. The meeting endorsed continued subgroup work to clarify overlap and to produce practical FAQs and guidance (see the Board’s endorsed document on MDR/IVDR interplay). The Commission’s Digital Omnibus proposal aims to make targeted adjustments across related digital legislation to reduce friction between instruments, while the Data Union Strategy addresses data governance interoperability essential for compliant AI development and evaluation.
International Alignment
Members reiterated the importance of international alignment on standards, evaluation methods and terminology to reduce fragmentation and to facilitate interoperability with trade partners. The Board noted ongoing dialogues with international partners and standardisation bodies, and acknowledged that EU interim guidance will need to be compatible with international standardisation developments. The Board also highlighted the strategic role of the AI Office in international cooperation and in promoting European approaches to trustworthy, interoperable AI.
Implementation Timeline
| Milestone | Target date / status |
|---|---|
| Sixth AI Board meeting | 4 December 2025 (held) |
| Digital Omnibus proposal published | 19 November 2025 (published) |
| Rotation handover (Chair) from Denmark to Cyprus | 1 January 2026 (planned) |
| Subgroup 2026 workplans implementation | 2026 (endorsed) |
| Interim Commission guidance (bridging standards) | Commission preparations announced at meeting; timelines subject to internal Commission work and Board consultation |
Sources and References
| Source | Type |
|---|---|
| Sixth AI Board meeting — European Commission (4 December 2025) | Primary Source |
| AI Board — European AI Office / European Commission | Primary Source |
| Digital Omnibus on AI Regulation Proposal — European Commission (19 November 2025) | Primary Source |
| Regulation (EU) 2024/1689 (Artificial Intelligence Act) — EUR-Lex | Primary Source |
Requirements for a company
What an organisation has to do under European Union - AI Act Implementation, at a glance. Not legal advice — the table below gives the provision and deadline for each item.
Must do
6- Maintain repositories for AI system test results and conformity documentation.Providers of AI systems
- Review Commission interim guidance when published.Providers of AI systems, Member States
- Consult guidance on the interplay between the AI Act and sectoral regimes like Medical Devices Regulation.Providers of AI systems in regulated sectors (e.g., health)
- Align national conformity assessment procedures.National competent authorities
- Implement incident reporting and market surveillance interoperability.National authorities, AI Office
- Ensure transparent appeal routes and align procedural safeguards in national enforcement mechanisms.Member States
Must not do
0Nothing in this category.
Should do
2- Consider using the General-Purpose AI Code of Practice as an adequate demonstration tool.Providers of General-Purpose AI systems
- Participate in regulatory sandboxes and interoperability pilots.Industry, research bodies, SMEs
Should not do
0Nothing in this category.
Who must do what
The obligations under European Union - AI Act Implementation, most serious first. Not legal advice — verify against the official text before relying on it.
| # | Who | Requirement | By when | Where | Severity |
|---|---|---|---|---|---|
| 1 | Providers of AI systems | Maintain repositories for AI system test results and conformity documentation. “practical instruments such as regulatory sandboxes and repositories for test results and conformity documentation.” | — | Implementation Framework | Important |
| 2 | Providers of AI systems, Member States | Review Commission interim guidance when published. “Review Commission interim guidance when published” | — | Compliance Checklist | Important |
| 3 | Providers of AI systems in regulated sectors (e.g., health) | Consult guidance on the interplay between the AI Act and sectoral regimes like Medical Devices Regulation. “The meeting endorsed continued subgroup work to clarify overlap and to produce practical FAQs and guidance (MDR/IVDR interplay).” | — | Relationship to Other Instruments | Important |
| 4 | National competent authorities | Align national conformity assessment procedures. “Align national conformity assessment procedures” | — | Compliance Checklist | Important |
| 5 | National authorities, AI Office | Implement incident reporting and market surveillance interoperability. “Implement incident reporting and market surveillance interoperability” | — | Compliance Checklist | Important |
| 6 | Member States | Ensure transparent appeal routes and align procedural safeguards in national enforcement mechanisms. “The Board addressed the need for transparent appeal routes and for Member States to align procedural safeguards in national enforcement mechanisms.” | — | Penalties, Liability, and Appeals | Important |
| 7 | Providers of General-Purpose AI systems | Consider using the General-Purpose AI Code of Practice as an adequate demonstration tool. “The Board considered voluntary tools like the General-Purpose AI (GPAI) Code of Practice as an adequate demonstration tool where applicable.” | — | Key Focus Areas | Recommended |
| 8 | Industry, research bodies, SMEs | Participate in regulatory sandboxes and interoperability pilots. “Participate in regulatory sandboxes and interoperability pilots” | — | Compliance Checklist | Recommended |
Related Regulations
Commission Decision of 24 January 2024 establishing the European Artificial Intelligence Office
European Union91% similar
Coordinated Plan on Artificial Intelligence — 2021 review
European Union90% similar
EDPB-EDPS Joint Opinion 1/2026 on the Proposal for a Regulation as regards the simplification of the implementation of harmonised rules on artificial intelligence (Digital Omnibus on AI)
European Union90% similar
White Paper on Artificial Intelligence: A European approach to excellence and trust
European Union90% similar
Communication: Fostering a European approach to Artificial Intelligence
European Union90% similar
© Regulations.AI · updated on 13-Jun-2026 · reviewed against official sources on 06-Sep-2026 using Gemini 3.6 Flash