The AI ​​Office is the central supervisory and enforcement body for the AI ​​Act. It was established by the European Commission in January 2024 and is located within the Directorate-General for Communications Networks, Content and Technology (DG CNECT).

Its overarching objective is twofold:

• Enforcement: To ensure that the rules of the AI ​​Act, particularly for the most powerful AI models, are applied and respected consistently across the Union.
• Promotion: To foster the development and use of trustworthy AI in Europe to strengthen innovation and competitiveness while protecting citizens’ fundamental rights.

The Office is intended to serve as a global reference center for a human-centered approach to AI.

The AI ​​Office’s responsibilities are wide-ranging and extend far beyond mere oversight.

Oversight of Advanced AI Models: Its primary task is the direct oversight of general-purpose AI models (GPAIs), particularly those classified as “systemic risk” (such as very large language models). The Office will monitor their compliance with regulations, initiate investigations into violations, and can impose sanctions.

Coordination and Harmonization: The Office serves as a liaison between the 27 EU Member States. It will coordinate national supervisory authorities to ensure a consistent interpretation and application of AI legislation across the EU. It will also support joint investigations in cross-border cases.

Development of Guidelines: To provide clarity for companies and developers, the AI ​​Office will develop detailed guidelines, implementing regulations, and standards for the practical application of the law.

Promoting Innovation: The Office will not only regulate but also foster an innovative ecosystem for AI “Made in Europe.” This includes supporting so-called “AI Factories” and promoting research and development in the field of trustworthy AI.

Support for SMEs: Small and medium-sized enterprises (SMEs) will receive targeted support. The Office will provide information platforms and resources to facilitate their compliance with the complex regulations.

International Representation: The AI ​​Office will represent the EU on the global stage and promote the European approach to trustworthy AI in international bodies and with partner countries.

To meet these diverse tasks, the AI ​​office is divided into several specialized units and will be staffed with approximately 140 experts, including technologists, lawyers, and economists.
… The structure includes, among other things:

• Regulation and Compliance Unit: The core team for monitoring and enforcing the rules, particularly for GPAI models.
• AI Safety Unit: Focuses on identifying and mitigating risks posed by AI systems.
• Excellence in AI and Robotics Unit: Responsible for promoting research, development, and the dissemination of AI in the EU.
• AI for Societal Benefit Unit: Focuses on the positive applications of AI, for example, in the health or environmental sectors.

The office will also work closely with the European AI Committee (in which the Member States are represented), a scientific panel of independent experts, and an advisory forum (which includes civil society and industry).