Key point: The KI-MIG establishes the national implementation framework for the European AI Act. The Federal Network Agency becomes the central supervisory authority with a new AI Market Control Chamber for high-risk systems. For the first time, there is a centralized complaint mechanism for citizens.
With the draft law to implement the European AI Act (KI-MIG), the German Federal Government creates the conditions for the first time for national implementation of EU regulation. The focus is on clear responsibilities, new supervisory structures, and the promotion of innovation.
The law to implement the Regulation on artificial intelligence (KI-MIG) transposes EU Regulation (EU) 2024/1689 – better known as the AI Act – into German law. The draft law is already available (Bundestag document 21/4594) and must now be debated in the Bundestag. The original implementation deadline has already passed on 2 August 2025. Entry into force cannot yet be foreseen.
The AI Act has been in force since August 2024 and will become largely directly applicable from 2 August 2026. The KI-MIG supplements the European requirements with national regulations, particularly where organizational and procedural structures are required.
The European regulation is based on a risk-based approach to AI systems, which provides for prohibitions on certain applications, requirements for high-risk AI, and transparency obligations. The German implementation law creates the institutional and procedural foundations for this.
Central innovations of the KI-MIG:
**Federal Network Agency as central AI supervisory authority**: The Federal Network Agency becomes the central market surveillance authority for AI systems with far-reaching enforcement powers. It is responsible for compliance with the AI Act to the extent that no special responsibilities apply, establishes a central point of contact for companies and citizens as well as for forwarding and coordinating complaints. Additionally, a coordination and competence center is being established.
**New supervisory structure for high-risk AI**: An independent AI Market Control Chamber is established within the Federal Network Agency to monitor high-risk AI systems, particularly in sensitive areas such as law enforcement, border control, justice, and democratic processes. At the same time, sectoral responsibilities remain, such as those of the Federal Financial Supervisory Authority (BaFin) in the financial sector or of state authorities for certain public applications. This results in a hybrid supervisory system of central control and sectoral expertise.
**Centralized complaint mechanism**: For the first time, a centralized complaint mechanism for AI violations is established. Citizens can report potential violations centrally, complaints are forwarded to the responsible authorities, and transparency regarding responsibilities is created.
Source: www.activemind.legal