Labelling requirement for AI content

Since the August 2, 2026 the transparency obligations from Article 50 of the AI Act apply. Not only for developers but for every company that uses AI. Anyone operating a chatbot on the website, publishing AI-generated images or texts, or using synthetic voices in customer service must make this recognizable for users. The European Commission has compiled exactly what needs to be labelled:

The AI Omnibus comes into force: Since July 27, 2026, the amendment regulation to the AI Act applies: The obligations for high-risk systems apply only from December 2027 or August 2028, simplified documentation and proof obligations apply for SMEs.

Who is affected by the AI regulation?

The regulation covers any company or organization that uses AI tools. Operators of AI systems are described in the AI Regulation as "natural or legal persons, authorities, institutions, or other entities that use an AI system under their own responsibility." This includes all professional applications. Only private, non-professional uses are exempt.

Thus, from February 2, 2025, Article 4 applies to all operators: sufficient AI competence of employees. How much further you need to go depends on what you use AI for. Take a closer look if you ...

• use AI in personnel selection, credit checks, or critical infrastructure – such applications are considered high-risk according to Annex III, even with purchased tools,

• operate in a regulated industry such as healthcare, financial services, or transportation,

• work with personal data or copyrighted content,

• develop or train your own AI systems – then you also apply as a provider, with significantly more extensive obligations.

The earlier you know the legal framework, the safer and quicker you can implement AI projects without having to make corrections later.

Implement AI Act with the Bavarian AI Innovation Accelerator

Training, workshops, and practical aids for implementing the AI Act. The Bavarian AI Innovation Accelerator is a project initially funded for two years by the Bavarian State Ministry for Digital Affairs. As part of the project, SMEs, startups, and the public sector in Bavaria are supported in complying with the European AI Regulation.

Implement AI Act: 5 steps for your company

The following approach has proven successful for many companies:

1. Step 1 – Check for impact: Clarify whether your company qualifies as an operator of an AI system within the meaning of the AI Regulation. This applies to any professional use of AI tools, regardless of company size or industry – only purely private use is exempt.

2. Step 2 – Carry out risk assessment: Assign your AI applications to the four risk levels of the AI Act, from harmless to unacceptable. This way, you can recognize which obligations apply specifically to you.

3. Step 3 – Ensure AI competence: Plan training measures for your employees in accordance with Article 4 of the AI Regulation – as an online course, face-to-face training, or self-study, suitable for the knowledge level and application area of your team. This not only meets the legal requirement: Trained employees recognize faulty AI results, use the systems more effectively, and assess more safely where use is sensible.

4. Step 4 – Obtain legal advice: Check especially in the case of self-developed AI systems as well as questions regarding data protection, copyright, and industry-specific requirements, whether you need legal support.

5. Step 5 – Regularly review the legal situation: Regulation is evolving dynamically. Follow current developments and continuously adapt your processes to remain compliant in the long term.

You can find more information on the AI Compass topic "Law and Regulation" in our download portal.