AI Is Not a Switch You Simply Flip

Everyone is talking about AI these days. In your view, what is being massively overestimated?

Florian Miller: The idea that you can simply buy an AI tool and automatically become more productive overnight. AI is not a switch you simply flip. The real effort lies in data, processes, and integration into existing workflows. Many companies focus on impressive demos but underestimate the key question: What problem is actually being solved? In the end, it is not the technology that matters, but whether development cycles are shortened, quality improves, or knowledge becomes more accessible.

How can companies tell whether an AI initiative is creating real value or is merely an innovation project without a business case?

Florian Miller: An AI initiative creates real value when it makes a measurable contribution to business objectives, for example through higher productivity, lower costs, improved quality, faster processes, or additional revenue. If success is measured only by the number of pilot projects, users, or technical features, without any demonstrable business benefit, it is more of an innovation project than a business initiative. If, after a year, all that exists is a prototype and a stack of presentation slides, it was probably an innovation project rather than a solution delivering business value.

What advice would you give companies?

Florian Miller: Focus on business impact rather than technology hype. Start with concrete use cases, establish clear responsibilities, invest in enabling and educating employees, and think about scaling from the beginning. The biggest success factor is usually not the technology itself, but the consistency and quality of execution across the organization.

What would you say to a CEO who has consciously decided against AI so far?

Florian Miller: Do not start with the question, “Do we need AI?” Instead, ask: “Where are we currently losing time, knowledge, or efficiency?” If there are clear answers to those questions, AI becomes worth exploring. Companies do not need to transform everything immediately. However, they should understand the opportunities that already exist today before their competitors take advantage of them.

Many companies fear that their data may end up in AI models. Is that concern justified?

Florian Miller: Yes, this concern should be taken seriously. Companies need to know exactly where data is processed, who has access to it, and which security mechanisms are in place. If these questions cannot be answered, sensitive data should not be processed using AI systems. At the same time, caution should not lead to paralysis. The key is to use AI in a controlled and responsible way.

Almost every company claims to use AI responsibly today. Why should customers believe them?

Florian Miller: Because responsibility is not created in a presentation; it is demonstrated in practice. When a company processes sensitive development data or operates in regulated industries, results must be traceable, systems secure, and decisions verifiable. That is exactly why we follow the principle of“AI amplified, human verified.” AI provides support, but responsibility remains with people. Our customers do not have to compromise on security or trust. These standards remain just as high as ever. At the same time, they can benefit from the efficiency and innovation potential of modern AI technologies. To make this approach visible, we launched our own AI initiative and developed an exclusive label for all AI-related services across our company.

“AI amplified, human verified” sounds logical, but what does it actually mean?

Florian Miller: This principle captures our philosophy perfectly. AI can analyze, structure, and generate recommendations. However, evaluation, interpretation, and accountability remain in human hands. This allows us to combine the speed of AI with human experience, expertise, and judgment. For us, this balance is the key to trustworthy AI.

What problem does the AI Stack solve that standard solutions cannot?

Florian Miller: Many standard solutions work extremely well for individual use cases. The challenge begins when organizations want to operate multiple AI solutions securely, at scale, and over the long term. This is where our AI Stack comes in. It provides a common foundation for security, governance, and data management. As a result, teams do not have to start every project from scratch and can turn AI ideas into productive solutions more quickly instead of creating isolated point solutions.

Can a label explain what a thousand presentation slides cannot?

Florian Miller: Perhaps not explain it, but make it easier to understand. For many people, AI is still an abstract concept. A strong label creates orientation and makes it visible what a technology stands for. And sometimes, that image stays in people’s minds longer than the hundredth presentation. The label visually represents capabilities such as knowledge access, automation, and intelligent assistance. The benefits for companies are tangible: less time spent searching for information, faster decision-making, and more efficient use of existing knowledge.

Our AI expert, Florian Miller

Florian Miller

Learn more about Artificial Intelligence at Bertrandt