EU AI Office
Brussels-based GPAI supervisor (DG CNECT)
Practitioner's note: This article is practice-oriented compliance documentation, not legal advice. We are a compliance specialist, not a law firm. For legally binding information please consult a licensed lawyer.
TL;DR
The EU AI Office (Article 64 EU AI Act) is the central EU body for GPAI models. It coordinates the Code of Practice, investigates GPAI providers (Article 92), imposes sanctions (Article 101), and advises the AI Scientific Board.
What is the EU AI Office?
Responsibilities:
- Maintaining the GPAI model register
- Coordinating the Code of Practice
- Investigations and sanctions for GPAI
- Advising the Commission and Member States
- International cooperation (US AISI, UK AI Safety Institute)
Relevance for SMEs: indirect — supervision of suppliers (OpenAI, Anthropic, Google). SMEs in their capacity as deployers are not directly addressed.
Practical example
The EU AI Office opens proceedings against GPAI provider X for inadequate training data documentation. Investigation: 12 months; sanction: EUR 8 million. Impact: rectification of the documentation within 6 months.
Frequently asked questions
Where is it based?
Brussels, with a target headcount of approximately 80 staff from 2026 onwards.
Who heads it?
Lucilla Sioli (Director of the AI Office) since 09/2024.
Are complaints possible?
Yes, the AI Office accepts complaints regarding GPAI models.