On 5 December 2025, an on-line workshop explored how electronic health record systems/ EHR systems users can best benefit from artificial intelligence (AI)-supported monitoring and decision support systems. It took place with the support of the Xt-EHR project. Over 60 people attended. Based around aspects of the European Health Data Space (EHDS) – but going beyond the space – the workshop captured the real requirements for meaningful interactions between EHR systems and AI solutions. 

What were the main learnings from the workshop

Contributors, Luc Nicolas (EHTEL), Zoltan Lantos, (Semmelweis University, Hungary), Dragan Sahpaski (SORSIX, North Macedonia) led off the workshop discussions. Their conversations covered:

Key messages

• “AI, data fluidity, and FAIRification is more prospective than what’s in the EHDS regulation.”
• “AI is vital to data fluidity.”
• “FAIR data is not enough on its own. We also need data fluidity.”
• “AI can actually make AI happen.”
• “Use good models, methods, and tools, and integrate them with AI tools.”
• “We want vendors to act as ‘bridges’ and not as ‘gatekeepers.’”

Main barriers

• Incomplete standardisation and semantic interoperability.
• Reliance on human coding in non-clinical silos.
• Problems with the quality and quantity of data.
• Testing is not continuous.
• Limited progress on data governance.
• Possible EHR-compliance could inhibit external innovation.

Recommendations

Seven potential recommendations focused on semantics, data quality assurance, user-centred design and continuous co-creation, strategic investments made at multiple levels, the balancing of transparency/explainability and ‘seamless AI’, going beyond the ‘still nascent’ aspect of AI algorithm integration, and the leveraging of AI for FAIRification.

⚜️ The meeting's report and videos are now available for EHTEL members, on the Member Workspace.

Main items of discussion

Input came from over a dozen attendees at the workshop. Their comments and experiences ranged over many practical challenges. Examples of challenges included the embedding of data in EHR systems; forerunner countries and regions handling data governance; the time-periods that new coding schemes take to be implemented; the provenance and tracking of data; the role of hospitals in the whole data-gathering/data-processing process; the importance of use cases (e.g., pre-diabetes/diabetes; cancer[s]); collaboration on the accuracy of data by people, patients, and doctors; the risks of AI misinterpretation of data coming from different institutions and countries; the importance of ‘guardrails’ for AI; the importance of skills, competences, and the alignment of training on AI alongside other EHDS-related training.

Other initiatives and European-financed projects that could provide useful messages included:

• CPME’s policy paper on the deployment of artificial intelligence in healthcare 
• EU-CiP 
• QUANTUM 

Key insights were:

• “The EHDS is like a digital Schengen area.”
• “The more data-sharing happens, the more errors will be spotted.”
• “There is a need for ‘essential performance’ [the maintenance of critical functions]. … The data just has to be better than a doctor at the end of their second shift.”
• “[Twin] a clinician who is ‘data-savvy with someone who is not.”
• “Do ‘rubber-ducking’ [debugging] with an AI assistant.”

Among the most challenging of questions exchanged among the attendees was how to make AI trustworthy and transparent at the same time as keeping its use discrete and not especially intrusive. How therefore to balance transparency/explainability and ‘seamless AI’?

What are the results of the workshop

As an outcome of the workshop, EHTEL is producing a working paper inspired by Xt-EHR WP5 deliverables which identifies “European-level requirements and constraints for FAIR data structures to enable communication between EHDS compliant EHR systems and algorithm-based tools” The report’s focus is on data, together with other legal, ethical and organisational aspects of data fluidity, FAIRification, the EHDS, compliant EHR systems. In a world that is ”EHR-connecting”, the report ends with a series of core recommendations on how to make practical progress on these critical issues.

How to offer more of your views

The working paper will be presented publicly in Helsinki on 21 January 2026, during the EHTEL Thought Leader Symposium, embedded in the Radical Health Festival Helsinki., wider communities of stakeholders are also encouraged to read and comment on the paper.

Come to the festival and Symposium to express your own views on EHR systems and AI. Join the ongoing debate!