05.09.2025

No Cloud, No Visibility: AI Needs The Cloud

In this interview, Felix Höger, CEO of EuroCloud, talks about the role of resilience in the cloud and AI environment. He explains why highly available infrastructures are so crucial for the operation of AI applications, why modern AI would be inconceivable without the cloud, and how companies can achieve a balance between cost-effectiveness, sovereignty and resilience when developing cloud-based AI strategies.

Mr Höger, when you look at current developments in the cloud and AI environment, what role does resilience play today, and why is it so crucial for the use of AI in particular?

Since EuroCloud was founded fifteen years ago, we have been promoting the idea that, in addition to agility, flexibility and scalability, resilience is also one of the advantages of the cloud. Thanks to redundant infrastructures, mechanisms such as geo-redundant failover and automated recovery, and virtually infinite scalability, workloads in the cloud can be operated without interruption. Ultimately, resilience describes nothing more than a system’s ability to deal with unforeseen events such as malfunctions or failures in such a way that functionality is not impaired, or only insignificantly so. In the world of AI, where applications must respond in real time and be continuously supplied with data, this ability is all the more important. The failure of an AI application can lead to significant financial losses, operational disruptions or security-related problems.

Why is it impossible to imagine AI without the cloud today – what makes cloud infrastructure so indispensable for AI solutions?

AI as a theoretical concept has been discussed since the 1950s. Without the cloud, however, the technology would never have found practical application in a broad range of companies, but would have remained limited to research institutes with high-performance computers. And AI solutions would not have been able to reach the level of maturity we see today. The cloud gives a large number of providers, especially start-ups, access to the computing resources they need to train or fine-tune AI models. As the EuroCloud association, we are very pleased that the cloud is now considered an indispensable enabler for further developments in the field of AI.

What advice would you give to companies that are developing AI strategies and aiming for cloud-based implementation – how can they remain sovereign and resilient? Do they have to weigh up economic efficiency and sovereignty?

In my discussions with application companies, I find that sovereignty is currently a dominant topic. However, sovereignty is not a binary state, but a process. Basically, companies that want to use AI applications in the cloud should establish clear governance from the outset: Which data is business-critical, which regulatory requirements apply, and which dependencies do you want to enter into or avoid? It is a matter of making a conscious decision about where to use proprietary services and where to use open standards. Then cost-effectiveness and sovereignty are not mutually exclusive. Those who choose the right cloud partners and rely on open interfaces can achieve both. In fact, there are great cloud and AI solutions, also developed by our member companies, that meet exactly these requirements in terms of openness and interoperability.

Felix Höger and the EuroCloud Executive Board cordially invite all eco members to this year’s EuroCloud Summit on 24 September 2025 in Dortmund to discuss this and many other exciting topics! Register now!

Ohne Wolke kein Durchblick: KI braucht Cloud