Artificial intelligence (AI) has long since made its way into everyday clinical practice. Following initial successes in the areas of diagnostics, documentation, and coding, the focus is now on integrating AI into hospital care in a safe, economically viable, and sustainable manner. The following expert commentary expands on the article by Uwe Heckert, published in the Krankenhaus KI Journal, with additional perspectives, practical examples, and concrete recommendations for action.
Many hospitals have already gained initial experience with AI. Whether automated documentation, medical coding, or image analysis—the results are promising. Yet individual pilot projects quickly reach their limits. Different systems, missing standards, and isolated data sets make widespread deployment difficult. The true added value only emerges when AI in hospitals is understood as a strategic capability: scalable, secure, and firmly integrated into existing care processes. Only in this way can innovations be sustainably established and operated economically.
Hospitals will in future not only use one or two AI applications, but operate numerous solutions in parallel—from the emergency department to the intensive care unit. This requires a shared technological foundation. Interoperable data platforms, FHIR (Fast Healthcare Interoperability Resources)-compliant interfaces, and cloud-capable architectures make it possible to introduce new applications more quickly and connect them securely with one another. Instead of isolated island solutions, a unified platform emerges on which innovations can be efficiently developed, operated, and continuously advanced. It is precisely this shift from individual projects towards an industrial AI architecture that determines long-term success.
A high-performance infrastructure is the prerequisite for unlocking the potential of AI in the hospital. Building on the Industrial AI Cloud of T-Systems, the vision of an AI Health Cloud is taking shape: a platform that securely consolidates health data, provides applications in a standardized manner, and makes innovations scalable. Hospitals benefit from central services, flexible provisioning of computing power, and an architecture that also supports future AI applications. At the same time, data protection, interoperability, and regulatory requirements are taken into account from the outset.
The necessary paradigm shift consists of no longer viewing AI as an experiment, but as an industrial capability—scalable, quality-assured, and firmly integrated into care processes.
Dr. Uwe Heckert, COO Health Industry, T-Systems & CEO Detecon International
The more AI is integrated into medical decisions within hospitals, the more important governance and trust become. Many applications are subject to the Medical Device Regulation (MDR) or must comply with requirements from the GDPR and, in future, the EU AI Act. Security, transparency, and quality should therefore be considered from the outset during development. Digital sovereignty is equally critical: hospitals must be able to understand at all times where health data is being processed, who has access to it, and how systems can be operated securely. Open standards and European cloud infrastructures provide the foundation for this.
The benefits of AI are already evident in everyday clinical practice. Automated documentation reduces administrative workload and creates more time for treatment. AI supports radiologists in evaluating medical images, summarizes extensive patient records, or assists in the preparation of tumor boards. In emergency departments, it can also structure information and accelerate decision-making processes. AI does not replace physicians—it expands their capabilities, improves the information base, and supports well-founded medical decisions.
The transition from initial AI projects to widespread adoption is not purely a technology decision. Successful hospitals combine medical expertise, modern IT, and clear governance. Equally important are interoperable data structures, quality-assured processes, and the acceptance of staff. Those who establish the right organizational and technical foundations today will be able to introduce new AI applications significantly faster in the future and operate them securely. In this way, AI in hospitals evolves from an individual project to a permanent component of future-ready healthcare.
The question is no longer whether AI will find its way into everyday hospital life, but how it can be used sustainably and responsibly. Individual pilot projects have already demonstrated the potential AI offers; now it is a matter of transferring these insights into a scalable infrastructure. Platforms, open standards, digital sovereignty, and strong governance form the foundation for this. Hospitals that set these course corrections today are creating the prerequisites for more efficient processes, better medical decisions, and future-proof patient care.
AI in hospitals refers to the use of artificial intelligence to support medical and administrative processes. AI can, for example, assist with diagnoses, automate medical documentation, analyze patient records, or prepare clinical decisions. The goal is to relieve the burden on specialists while simultaneously improving the quality of care. Responsibility for medical decisions always remains with the treating staff.
AI already supports many hospitals today, including in radiology, pathology, and medical documentation. Further areas of application include coding, the analysis of large volumes of data, the summarization of patient information, and support in emergency departments or tumor boards. With a suitable platform, these applications can be gradually expanded and interconnected.
A successful AI strategy requires more than high-performance algorithms. What is decisive is an interoperable IT landscape, high-quality data, open standards such as Fast Healthcare Interoperability Resources (FHIR), secure cloud infrastructures, and clear governance rules. Equally important are the early involvement of staff and the consistent integration of applications into existing clinical processes.
Health data is among the most sensitive information of all. This is why AI solutions in healthcare must meet the highest requirements for data protection, information security, and transparency. European cloud infrastructures, digital sovereignty, encryption, and compliance with regulatory requirements such as GDPR, MDR, and the EU AI Act create the foundation for a trustworthy use of AI in hospitals.
Pilot projects demonstrate the potential of individual applications, but are often difficult to transfer to other areas. Only a shared platform with standardized interfaces, central governance, and scalable infrastructure makes it possible to operate multiple AI applications efficiently, securely, and economically throughout the entire hospital.
Cloud platforms provide the necessary computing power and scalability for modern AI applications. At the same time, they facilitate the integration of new services, secure data exchange, and the operation of high-performance models. For hospitals, sovereign cloud solutions are becoming increasingly important, as they combine data protection, compliance, and digital sovereignty with the requirements of modern AI.
