AI in Business? Transparency and Labeling Duties under the EU AI Act

Chatbots, automated texts, AI-assisted research, image generators and internal assistant systems: AI tools have become a regular part of everyday business operations.

Many companies use artificial intelligence to make processes more efficient, work faster and create content automatically.

Although the AI Act primarily addresses high-risk AI systems, even everyday AI applications in a professional context may trigger transparency obligations. What matters is not only the risk category, but above all the specific use case.

📅 This will become particularly relevant from 2 August 2026, when the transparency obligations under Article 50 AI Act become practically effective.

Numerous everyday applications are already affected today, such as:

• AI tools for internal use, customer support, HR, writing, visuals and content creation

• Integrated AI in software-as-a-service applications

• AI for the analysis, structuring and evaluation of press, social media, audio or video content

• AI-supported tutor and assistance systems

• AI for improving rankings and recommendation systems for platforms

  
  

A company may use AI with only low or minimal risk and still be required to inform users about the use of AI or to disclose AI-generated content.

Five Steps Toward Compliant AI Use ✅

We recommend five key measures to help companies assess whether transparency and compliance obligations under the EU AI Act may arise.

🔎 1. Audit AI Tools

AI systems are often a factual “black box”. Content may be stored, analyzed or even used to train external models without it being fully transparent how the data is processed. Companies should record all AI tools used internally and document in which product they are used, for what purpose, and whether data is stored, analyzed or used for the training of external models.

📚 2. Review Contracts

Existing contracts with clients, service providers and employees should be reviewed. How is the use of AI regulated? Which transparency, documentation and labeling duties must be implemented, and what information must the external AI provider supply for this purpose?

🎓 3. Train Teams

Under Article 4 AI Act, all staff should have a sufficient level of AI literacy. This requires awareness of how to use AI systems responsibly, including clear rules on the use of personal or confidential data in connection with the GDPR and the AI Act.

📝 4. Internal AI Guidelines and AI Charter

An AI Charter sets out the company’s principles and commitments for the responsible, transparent and secure use of AI. Internal AI guidelines, by contrast, should provide concrete organizational and legal rules for employees on the practical use of AI systems in day-to-day business.

👀 5. Editorial Control

AI-generated content, decisions or analyses should be reviewed and confirmed by natural persons before being used or published.

🗣️Using AI as a Tool for Transparency and Communication

AI should not be seen only as an efficiency tool. A transparent, structured and responsible approach to AI builds trust and makes companies not only more credible, but also future-ready.

How Mars-IP Supports Companies 🚀

Mars-IP assists companies with the practical and legally sound implementation of the requirements of the EU AI Act: from classifying existing AI applications and assessing transparency and documentation duties to drafting internal AI guidelines and building appropriate compliance structures.

Contact us

Image: ChatGPT