EU AI Act: Transparency as an opportunity

With the new transparency requirements of the EU AI Act taking effect on August 2, 2026, one key question comes into focus: How can companies ensure the responsible use of AI? For Christoph Bräunlich, Managing Director of AI at BSI Software, one thing is clear: Transparency is far more than a regulatory requirement – it is becoming a key distinguishing feature. In this interview, he discusses ethics, self-determination and the future of “human-AI collaboration.”

When companies disclose when and how AI is used, they offer their customers guidance and reassurance. This strengthens a brand’s credibility and can become a true distinguishing feature. If you want to have satisfied customers in the long term, there’s no getting around trustworthy AI. 

For me, AI is trustworthy when it aligns with our values as outlined in our Code of Conduct. This includes preventing harm, ensuring fairness and enabling people’s self-determination. Transparency and accountability are equally important: Companies must make it clear how AI is used, critically examine its impacts and actively minimize risks – just as BSI Software does when developing AI systems. Trust is not built by technology alone, but through its responsible use.

Trustworthy AI puts people at the center. It respects data protection, avoids discrimination and leaves room for human oversight. Europe, in particular, brings a unique perspective to this: We have traditionally placed greater emphasis on values such as data protection, governance and accountability. At a time when trust in digital technologies is becoming increasingly important, this very approach can become a decisive competitive advantage.

Transparency means not leaving people in the dark about when they are interacting with AI or consuming AI-generated content. This could involve labeling an AI-generated image, indicating that customer service uses AI agents or disclosing that a message was generated automatically. Transparency does not mean explaining every technical detail. It’s about giving people the information they need to act deliberately and autonomously. For example, if a customer is on hold, she should be able to decide for herself whether she wants to wait for personal assistance or use the quick support provided by an AI agent.

Yes, of course, because customers should know who or what they’re interacting with so they can make an informed decision about how they’d like to communicate – as the example with the hold queue shows. That doesn’t mean AI is less valuable than human interaction. On the contrary: Many end customers appreciate the fast and fuss-free support provided by AI agents. In the future, people will increasingly decide for themselves when they prefer the efficiency of an AI agent and when they deliberately seek personal contact. This freedom of choice will be an important part of a good customer experience.

Absolutely. Compliance with legal requirements is the bare minimum; trust must be earned beyond that. An AI application can meet all legal requirements and still be perceived as non-transparent, manipulative or unfair. Responsible companies such as BSI Software therefore ask not only, “Are we allowed to do this?” but also, “Should we do this?” For us, the EU AI Act is not an obstacle but an opportunity. It creates space to develop AI applications in a way that not only minimizes risks but also creates real added value through transparency, accountability and digital ethics. That is precisely where the ethical dimension of “AI ready” begins.

AI will increasingly take on tasks, speed up processes and lighten the workload for employees. At the same time, human skills such as empathy and responsibility are becoming more important. The future, therefore, does not lie in a “human versus AI” scenario, but rather in intelligent collaboration within hybrid teams. AI should support humans, not replace them. Humans must set the direction, bear responsibility and make the final decision. That is why we speak of “human in the lead” and “human in the loop.” Both roles are important and enable co-creation and co-working, respectively.