Wie wird der EU AI Act international durchgesetzt?
The EU AI Act is the world’s first comprehensive law regulating artificial intelligence. Many companies, developers, and users are asking: Does this apply to me? And what happens if the rules are broken? In this article, we answer the most important questions about the AI regulation, its international enforcement, and what AI Act compliance means in practice.
What is the EU AI Act and who does it apply to?
The EU AI Act classifies and regulates AI systems according to their level of risk. It applies to all providers, operators, and users of AI systems that are active in the EU or whose systems are deployed there — regardless of where the company is based.
The law distinguishes four risk categories: unacceptable risk, high risk, limited risk, and minimal risk. Systems with unacceptable risk, such as social scoring by government authorities, are prohibited. High-risk systems — for example, AI used in recruitment or healthcare — are subject to strict requirements regarding transparency, documentation, and human oversight.
How is the EU AI Act enforced outside the EU?
Every company that offers or deploys AI systems in the EU must meet the requirements of the AI regulation. Companies from third countries without an EU establishment must appoint an authorised representative in the EU. Companies outside the EU that distribute or operate AI systems in Europe must, among other things:
- appoint an authorised representative in the EU
- provide technical documentation and proof of conformity
- carry out a conformity assessment for high-risk systems
- grant authorities access to relevant documents upon request
Which authorities are responsible for enforcement?
Enforcement takes place on two levels: The AI Office coordinates at EU level and is responsible for high-risk systems of EU-wide relevance as well as prohibited AI practices. At national level, member states establish their own market surveillance authorities. In Germany, the Bundesnetzagentur and the BSI are considered likely candidates.
What happens if the EU AI Act is violated?
Violations can result in substantial fines: the use of prohibited AI systems can be penalised with up to 35 million euros or 7 percent of global annual turnover. For other violations, fines of up to 15 million euros or 3 percent of turnover are provided for. In addition, authorities can order the recall or discontinuation of an AI system. Lower caps may apply for SMEs.
How can companies demonstrate AI Act compliance?
AI Act compliance can be demonstrated through technical documentation, conformity assessments, and internal governance structures. Specifically, companies should:
- classify AI systems by risk category
- create and keep technical documentation up to date
- fulfil transparency obligations towards users
- implement human oversight mechanisms
- define internal responsibilities for AI governance
- register systems in the EU database for high-risk AI
Anyone who engages early with AI competence and AI Act training has a clear advantage — because compliance starts with understanding the applicable requirements.
When do the obligations of the EU AI Act take effect?
The EU AI Act comes into force in stages. Published in August 2024, prohibited AI practices are deemed unlawful from February 2025. Requirements for high-risk systems apply from August 2026, with exceptions until August 2027.
How we support you with AI Act compliance
We at mITSM help you in a structured and practical way. Our offering includes:
- training courses on the EU AI Act for various roles within your organisation
- training on AI fundamentals and responsible AI use
- certifications through ICO-Cert as a recognised certification partner
- in-house training tailored to your specific company situation
- flexible formats as classroom training, online live training, or e-learning
Take a look at our training offering in the area of AI competence and get in touch with us.
Dieser Inhalt wurde mithilfe von KI erstellt und kann Fehler enthalten.