MAC IT-Solutions at the K5 2026: AI meets ERP - live in Berlin
Home / MAC IT-Solutions at the K5 2026: AI meets
C. Minor
28 May 2026
Regulated industries can use AI and the cloud safely if they do not introduce new technologies in isolation, but combine them with clear responsibilities, controlled data processes and traceable decisions from the outset.
This is creating a new form of digital transformation, particularly in pharma, biotech, medical technology and animal health. It is no longer just about making modern technologies available more quickly. The decisive factor is whether companies can shape innovation in such a way that it remains sustainable in the long term even under regulatory requirements - GxP, EU AI Act, FDA 21 CFR Part 11.
AI and the cloud promise speed, automation and new ways of handling data. At the same time, they increase the requirements for transparency, documentation and control. Anyone who considers these two perspectives separately risks new complexity. Thinking about them together creates the basis for a resilient digital architecture.
Many companies start with AI via individual use cases: an assistant for documents, automatic classification, a forecast in the quality environment or a tool for analysing large amounts of data. This is understandable because the benefits should be quickly visible.
In a regulated environment, however, this view is not enough. An AI application is rarely just a single tool. It utilises data, generates results, influences decisions and is often linked to existing systems. This automatically creates a link to processes, responsibilities and regulatory evidence.
Companies should therefore clarify this early on:
The difference between an experiment and a resilient AI deployment does not lie in the technology alone. It lies in the ability to bring together benefit, risk and responsibility in a clean way.
Traditional software follows defined rules. AI systems work differently: they are based on data, training levels, probabilities and models that cannot always be explained linearly. This is precisely what makes them particularly challenging in a regulated environment.
In practice, Explainable AI (XAI) does not mean that every technical detail of a model has to be explained in a generally understandable way. However, companies must be able to demonstrate why an AI system is used in a specific context, what data it is based on and how results are verified.
Particularly relevant here are:
Responsible AI: Clear rules for ethically acceptable, non-discriminatory modelling results
Explainable AI: Traceability of modelling decisions for auditors and specialist departments
Human Oversight: Defined checkpoints at which people evaluate and approve results
Training and test data: Complete documentation of origin, version and quality
Model and system changes: Audit-proof change management analogue to Computer System Validation (CSV)
Limits of automated decisions: Clear definition of which decisions always require human approval
In a regulated environment, AI must not become a black box in the process. It must be embedded in such a way that people can understand, evaluate and take responsibility for decisions.
Cloud and SaaS solutions are changing the way business systems are operated. Infrastructure, technical maintenance and certain security mechanisms lie more with the provider. However, the regulatory responsibility for the company's own processes, data and evidence remains with the company.
This is particularly relevant because modern cloud platforms are constantly evolving. New functions, updates and technical changes occur more frequently than in traditional on-premises environments. For regulated companies, this requires a different operating model - no higher risk, but more governance.
Instead of viewing the cloud merely as an IT infrastructure, pharmaceutical companies and medical technology manufacturers should see it as part of their compliance model. This includes:
The cloud can make companies faster, more scalable and more flexible. But it does not exempt them from ensuring the validated and controlled status of their business-critical processes.
Many compliance models originate from a world in which systems remained stable over a long period of time. A project was implemented, validated and then further developed using defined change processes. AI and cloud landscapes work more dynamically.
Data changes. Models are customised. Systems receive regular updates. This makes compliance less of a one-off project phase and more of a permanent operational task - continuous compliance instead of one-off validation.
In practice, this means
The decisive factor is whether companies can operate technology in a controlled manner in the long term - not just whether they have introduced it correctly.
Modern governance for AI and the cloud should not be seen as an additional administrative burden. It is the framework that makes innovation scalable in the first place.
Models that are too rigid slow down innovation. Models that are too loose jeopardise traceability and compliance. Successful companies create guard rails within which new technologies can be used safely.
The result is not a control apparatus for the sake of control. The result is an operating model that combines speed and regulatory security.
The EU AI Act creates a binding legal framework for artificial intelligence in Europe for the first time. This is particularly relevant for regulated industries: AI applications in medical devices, pharmaceutical production or clinical decision support systems can be categorised as high-risk AI under Article 6 - with corresponding requirements for transparency, documentation, human oversight and technical robustness.
For companies, this means that AI strategies not only need technical architecture, but also organisational maturity. Anyone who introduces AI today without considering the regulatory dimension risks subsequent requalifications and compliance gaps.
Market analyses also underpin this correlation. The Microsoft Dynamics 365 Total Economic Impact™ report by Forrester shows that standardised processes, greater transparency and reduced manual effort generate considerable economic effects - and that these very factors also strengthen compliance and efficiency in a regulated environment.
Explainable AI (XAI) refers to the ability of an AI system to present its decision-making principles in a way that is comprehensible to humans. In a regulated environment, this means that companies must document why an AI model is used, what data it utilises and how results are verified - comparable to the verification requirements under EU GMP Annex 11.
The EU AI Act categorises AI systems according to risk classes. AI in medical devices, clinical decision-making systems or critical infrastructure is often considered high-risk AI (Art. 6). These systems are subject to obligations regarding transparency, technical documentation, human oversight and conformity assessment - much of which overlaps with existing GxP requirements.
Not mandatory, but it must be assessed. Risk-based change management means that every update is checked for its impact on GxP-relevant processes. If the impact is low, a documented risk assessment is sufficient. If the update affects validated processes, a targeted requalification step is required.
AI and the cloud are fundamentally changing regulated companies. They open up new opportunities to automate processes, make knowledge more usable and make more data-based decisions. At the same time, they are increasing the requirements for responsibility, traceability and control.
Successful deployment is therefore not the result of individual tools, but of a clear interplay of technology, data quality, governance and operating model.
Compliance is not the opposite of innovation. Properly understood, it is its prerequisite: it ensures that AI remains permanently usable, auditable and accountable in the regulated environment - and thus becomes scalable in the first place.
Home / MAC IT-Solutions at the K5 2026: AI meets
Home / collana health expands expertise in the Swiss healthcare sector With
collana is once again supporting the Deutschlandstipendium scholarship at Europa-Universität Flensburg in the 2025/26 funding year. In this way, we are supporting talented and high-performing students from the STEM subjects and specifically strengthening young talent in our region.