AI is meant to make processes faster. At the same time a new duty is approaching: transparency about when content comes from AI. For SMEs this is mainly a question of the right order. Whoever thinks both together avoids expensive rework later.
What changes from August 2026 under the EU AI Act?
From 2 August 2026 the transparency obligations of the EU AI Act take effect. Providers of generative AI must mark their outputs, whether text, image, audio or video, in a machine-readable way as artificially generated. Anyone who uses AI to create deepfakes must disclose that the content is artificial, even without intent to deceive. Violations can be fined up to 15 million euros or three percent of global annual turnover. For Germany an interim visual label with the abbreviation "KI" is under discussion.
Why is transparency also operationally useful?
Transparency is a duty and at the same time quality control. Whoever documents where AI sits in a process and who is responsible for the result keeps control over risk and error. This fits a lesson from practice: a widely cited 2025 MIT study found that around 95 percent of enterprise AI pilots deliver no measurable effect on the bottom line, often because it stays unclear where the AI actually works. Clean labelling forces exactly that clarity.
How do SMEs combine efficiency and compliance?
Pragmatically, not anxiously. We treat transparency as a fixed part of the process, not an afterthought. Concretely: for each AI use case, record which content is generated, who approves it and how it is labelled. Efficiency stays intact and the obligations are met before they become a problem. Compliance becomes a side effect of good processes rather than a brake.
Further reading
These articles help with the next decision.
- Understanding AI Agents: Practical Insights
- Why AI Understanding Often Fails in SMEs
- Why Medium-Sized Businesses Choose the Wrong AI Tools
When do the EU AI transparency rules apply?
The transparency obligations of the EU AI Act apply from 2 August 2026. They cover labelling of AI-generated content and disclosure of deepfakes.
Must AI content be labelled?
Yes. Generative AI outputs must be marked machine-readable as artificial, and deepfakes disclosed openly, even without intent to deceive.
What penalties apply for violations?
Fines of up to 15 million euros or three percent of global annual turnover, whichever is higher.
How do you combine efficiency and compliance?
By making transparency a fixed part of each use case: record what content is generated, who approves it and how it is labelled. Compliance then follows from good processes.