The EU AI Act: The new era of AI regulation in Europe

With the EU AI Act, the European Union has passed the world's first comprehensive legislation to regulate artificial intelligence (AI). The law came into force on August 1, 2024, and the first bans and obligations for companies and authorities have been in force since February 2025. The aim is to make AI technologies safer and more transparent while enabling innovation.

In this article, we explain what the EU AI Act means, what new regulations companies must observe and why mandatory training plays a key role.

The EU AI Act at a glance

The AI Act divides AI systems into four risk classes:

1. Unacceptable risk - Prohibited:

   • Social scoring by authorities or companies

   • Emotion recognition at the workplace or in educational institutions

   • Real-time biometrics for identification in public spaces

2. High risk - strict regulations:

   • AI in sensitive areas such as medicine, finance, law enforcement or education

   • Mandatory risk assessment, documentation and human oversight

3. Limited risk - transparency obligations:

   • Chatbots and deepfakes must be labeled as AI-generated

4. Minimal risk - no regulation:

   • AI systems in video games or spam filters

This categorization is intended to ensure that AI with a high social impact must meet particularly strict requirements, while low-risk innovations are not unnecessarily hindered.

New regulations from February 2025

The first specific bans on certain AI applications have been in force since February 2, 2025:

Strict ban:

• Emotion recognition in the workplace and in educational institutions

• Biometric categorization by ethnicity, religion or political persuasion

• Social scoring by companies or authorities

• Manipulative AI that pushes users to make certain decisions

Stricter sanctions:

• From August 2025, penalties of up to €35 million or 7% of annual global turnover will be imposed for breaches of the law.

New compliance obligations:

• Companies must systematically assess and document AI risks.

• High-risk AI systems must be regularly checked and certified.

• Users must be informed about the use and functioning of AI.

These measures make the EU AI Act one of the strictest AI regulations in the world, comparable to the GDPR in the area of data protection.

What does the AI Act mean for companies?

For companies, the EU AI Act primarily means

Carry out risk assessments: All AI systems used must be assessed according to their risk level.
Comply with documentation obligations: Transparent algorithms, comprehensible decision-making processes and data sources must be recorded in detail.
Implement mandatory training: Companies must ensure that employees who use or manage AI are adequately trained.

The training obligation in particular is often underestimated, but is crucial to avoid fines and legal risks.

Mandatory training: why companies need to act now

According to the EU AI Act, all employees who work with AI systems must have sufficient knowledge of their opportunities, risks and regulatory requirements. In concrete terms, this means

• Understanding AI risk classes and compliance requirements

• Training on transparency obligations and ethical standards

• Confidence in dealing with AI-supported decisions

Ohne entsprechende Schulungen können Mitarbeiter unbewusst gegen Vorschriften verstoßen – und das Unternehmen riskiert hohe Strafen.

💡 Our recommendation: Provide your employees with targeted training in order to comply with legal requirements and minimize liability risks.

👉 Our training courses on the EU AI Act

🎓 Reading tip: In our follow-up article, we go into detail about the training obligation in the AI Act - read more here! Find out more about the training obligation.

Conclusion: The EU AI Act will change the way AI is handled in the long term

The EU AI Act is a milestone in AI regulation that presents companies with new challenges. While it is not intended to block innovation, it provides clear rules and transparency in the use of artificial intelligence.

Companies should prepare at an early stage to avoid legal risks and secure competitive advantages. In particular, the obligation to provide training is a key aspect - because only those who use AI safely and in accordance with the rules can be successful in the long term.

Seize the opportunity and invest in AI expertise - we will support you!