Interview with Jean-Philippe Arié

Healthcare Week Luxembourg 2026 is built around an ambitious guiding theme: Resilience of Tomorrow’s Healthcare ,From Preparedness to Antifragility. Two days to connect People, Systems & AI ,not as separate tracks, but as one equation that must hold when everything accelerates.

For Jean-Philippe Arié, the starting point is clear: in 2026, the most useful role of innovation is not to impress ,it is to strengthen the foundations.

“In 2026, the most useful role of innovation is to strengthen the base that makes the system resilient: interoperable, secure and governed tools that ensure continuity of care and reliable information, even in times of crisis. We are not looking for the most brilliant technology, but for what enables us to stand the test of time and maintain trust through standards.”

This pragmatic stance echoes the views of other members of the Scientific Committee. Sylvain Vitali, Director of the Luxembourg Hospital Federation, states: “AI must increase the resilience of the system, not create new critical dependencies.” And Marianna Bevova, strategic advisor at the University of Luxembourg, reminds us that “in moments of tension, technology does not compensate for a weak culture ,it exposes it.”

Jean-Philippe Arié goes further by identifying the speed of change itself as a systemic challenge. Digital innovation cycles are accelerating, and AI is a catalyst for this ,provided that data can actually be used.

“Thinking that we can install a system tomorrow for the next 10 years seems to me to be a common mistake. Today’s new AIs are being restrained to give us time to transform, but that will not always be the case.”

The implication is significant: resilience is not about building once and then maintaining. It is about the system’s ability to adapt to each new iteration of innovation. And this raises a deeply human question: “For staff, having a technological update every morning will not necessarily always be easy to live with. We will need to support the interfaces between technological innovation and people. During crises, technology will adapt very quickly thanks to anticipated scenarios ,but what about humans?”

“Since March 2024, the new European EHDS regulation has imposed interoperability standards and governance principles aimed at ensuring transparency and trust in the exchange of medical data. The entire community ,doctors, pharmacists, hospitals, innovators and patients ,is entering a new paradigm that will enable the development of tomorrow’s data-driven medicine.”

He also makes a direct link with crisis management:

“These different regulations introduced in Europe make it possible to better structure data, from its creation through to its use. In the event of a crisis, Europe will be able to use reliable and sovereign data to make decisions. In other regions, data will be less structured, less validated and less well governed, which may lead to decisions potentially based on AI hallucinations.”

Innovation projects often succeed at the pilot stage but fail to scale. Jean-Philippe Arié identifies the core issue: “The main wall to break down is not technology ,it is the alignment between governance, regulatory frameworks and operational responsibilities.”

Pilots work because they are carried by motivated teams, within a controlled and temporarily exceptional framework, with the best experts. Scaling up, especially in a cross-border context, requires clear rules on data, responsibilities, interoperability and decision-making.

“As long as these issues are not addressed collectively, we will continue to succeed with pilots… without succeeding in their deployment or in the market adoption of solutions.”

This resonates with Guillaume Schaack’s observation that the ambition of HWL is precisely to “transform shared observations into dynamics of cooperation” ,and with Dr Martine Goergen’s call to “guarantee hospitals the means to prepare and meet the requirements of critical infrastructures.”

In degraded situations, Jean-Philippe Arié believes that robustness is not simply about a tool continuing to produce the same results.

“In times of crisis, the user must also become more critical. To do so, they need to understand the basic functioning of the tool in order to interpret the results themselves.”

He draws a clear line between certified medical AI ,whose functioning is known, tested and explainable ,and consumer generative AI, where the user does not understand how the tool reached a certain conclusion and may not realize when a result is misleading.

“Digital medical devices, including medical AI, are tested and their functioning is known, which is reassuring in terms of robustness. Medical staff are very well trained to be alert to unusual situations. This is why the innovation–human pairing, and training, are so important.”

Finally, when asked about the most common trap in innovation serving resilience, Jean-Philippe Arié is direct:  “We still see too many projects focusing on highly sophisticated tools, while the fundamentals are not in place: heterogeneous data, little or no interoperability, and poorly defined uses. In degraded situations, these solutions are the first to collapse, disrupting the work of healthcare professionals.”

Jean-Philippe Arié’s message for HWL26 is both pragmatic and visionary: innovation that strengthens resilience is not the most spectacular ,it is the most interoperable, the best governed and the most human-centred. It is deployed through standards, sustained through governance and improved by the people who use it.

Exhibit • Participate • Connect: www.hwl.lu