HLTH Europe 2026: What Amsterdam Revealed About the Next Phase of European Digital Health
Four days. Five thousand attendees from fifty-plus countries. RAI Amsterdam. And then it ends, and you have to ask: what did the event actually say?
Not the programme. Not the speaker list. The collective conversation that happens when a thousand C-suite leaders, a hundred regulators, and two hundred startups are in the same building for four days, trying to make sense of a sector moving faster than any single organisation can track.
Before HLTH Europe opened on June 15, I published a piece reading five structural signals from the agenda: AI moving from zone to infrastructure, interoperability as a permanent fixture, the Health Transformation Summit format as a diagnostic instrument, pharma as a structural partner rather than a guest track, and the shift of care beyond hospital walls.
Reading a conference programme as a market signal rather than a schedule is a method I find consistently useful. Agenda writers reflect where their audience genuinely is; they cannot afford to get this wrong. The question after four days is which of those signals held, and where Amsterdam pushed further than the programme alone had suggested.
AI Measurement Replaced AI Deployment as the Dominant Conversation
The AI@HLTH zone was central to the programme from the first day. What moved faster than expected was how quickly the sessions left deployment questions behind.
The room was not interested in “should we adopt AI?” The WHO/Europe data from April found 74% of European hospitals using AI in diagnostic workflows. That has settled the question for the majority of organisations present at RAI. The conversation that filled the dedicated sessions and the corridor exchanges was harder: how do you govern AI systems operating at scale, how do you attribute outcomes to specific model decisions, and how do you build vendor relationships where performance is the deliverable rather than the pitch?
This shift from deployment to measurement is not semantic. It changes procurement conversations, board discussions, and vendor relationships across the sector. A hospital system that has signed a five-year AI contract without measurement architecture embedded in it has a different problem in 2026 than it had in 2023. Amsterdam spent four days surfacing that problem. The organisations with answers were visibly more confident in every room.
The EHDS Debate at the Centre of RAI
“450 days of EHDS: Can it still give Europe an edge?” brought together the Swedish e-Health Agency, the French Health Data Hub, the Polish Ministry of Health, and Techniker Krankenkasse. The debate was not about whether EHDS matters. It was about whether the implementation pace across member states matches the ambition of the regulation.
Countries with more mature national digital health infrastructure are already moving into operational planning. Countries with more fragmented systems are still working through baseline compliance mapping. The first Implementing Act, published in April 2026, has sharpened the timeline considerably. Member states that treated EHDS as a future regulatory exercise are discovering it is a present infrastructure project. The distance between those who understood this in 2024 and those who understood it in 2026 is already measurable.
The milestones that matter most for secondary data use and cross-border health data access extend toward March 2027 and beyond. In the context of data governance cycles, procurement timelines, and the IT project schedules of large health systems, that is closer than the date suggests.
Cybersecurity Was Promoted, Not Just Present
In previous editions of HLTH Europe, cybersecurity appeared in the programme as a serious but bounded topic, the necessary addition alongside the more forward-looking tracks. In 2026, it carried a different weight, and the conversations in those sessions reflected why.
The sector has processed enough incidents: ransomware against hospital systems, AI-driven intrusions documented in late 2025, patient data breaches that reached national media. The sector has moved past the abstract risk framing. The conversation at HLTH Europe 2026 was operational: penetration testing frequency, vendor resilience requirements embedded in procurement contracts, the relationship between MDR compliance documentation and EU AI Act data governance obligations.
The EU AI Act transparency obligations for conversational and synthetic content AI systems come into force on August 2, 2026, six weeks from now. The compliance requirements for high-risk standalone AI systems listed under Annex III follow in December 2027. For healthcare organisations with AI tools already in clinical workflows, this is not a future planning horizon. It is a current operational one.
The organisations treating cybersecurity and regulatory compliance as infrastructure functions rather than annual compliance exercises were running a different conversation at RAI than those that were not.
The Pharma Presence That Changed Character
The Global Pharma Summit ran across both days of the Health Transformation Summit, and the conversations there reflected a vocabulary shift worth noting.
The framing in previous years was partnership: pharma working with digital health companies to improve clinical trial access, patient recruitment, or post-market surveillance. The framing at HLTH Europe 2026 was more specific. Pharma is looking for data governance partners, and the distinction from technology partners matters considerably.
A technology partner delivers a tool. A data governance partner builds the infrastructure through which real-world evidence is generated, validated, and used in regulatory and commercial decisions. The convergence of GLP-1 drugs reshaping metabolic care, companion diagnostics tied to oncology AI pipelines, and the Cancer Image Europe initiative (targeting 60 million cancer images in accessible format by end of 2026) is producing a new category of asset: structured, longitudinal clinical data at continental scale.
European health organisations that can demonstrate credible data governance frameworks and EHDS-ready infrastructure are entering pharma conversations at a different level than those leading with their AI stack alone. Amsterdam made this visible across multiple sessions.
What the Funding Data Confirmed
The investment environment that HLTH Europe reflected was already visible in the Q1 2026 numbers before the event opened. Total digital health funding in Europe reached $1.2 billion in the first quarter, a 44% decline in capital volume and a 46% drop in active deal count compared to the same period a year earlier. The average VC deal size rose 8% to $21.1 million.
This is not a contraction. It is a concentration. Capital is moving toward larger, later-stage platforms that can demonstrate clinical evidence, unit economics, and regulatory compliance architecture. Standalone tools without those properties are facing a shakeout that the HLTH Europe startup programming was honest about.
The “Regulatory Darwinism” filter combines EU MDR certification costs (between 200,000 and 600,000 euros per device) with EU AI Act high-risk compliance requirements. The combination is functioning as a market selector. Platforms that have navigated it are commanding premium valuations. Those that have not are running out of runway.
Three Things to Take Into the Second Half of 2026
EHDS is an organisational transformation project, not a compliance checklist. The organisations that treated the April 2026 Implementing Act as a prompt to begin actual work are already ahead of those still in observation mode. The 2027 and 2029 milestones are far enough away to feel abstract. Build the data governance function now, or inherit the compliance cost under deadline pressure later.
AI value is measured in outcomes, not in deployment rate. Hospital CIOs and institutional investors are reading from the same playbook in 2026. Platforms that deliver documented, measurable clinical ROI (reduced administrative time, improved diagnostic throughput, attributable patient outcomes) are capturing the capital and the contracts. The deployment-without-measurement model has a shortening shelf life. (See also: Why 70% of providers can’t measure what their AI does.)
The pharma/digital health intersection now requires governance expertise. The next commercial frontier in this space is not AI-powered drug discovery as a concept. It is the practical architecture of real-world evidence generation: how structured clinical data flows into regulatory-grade evidence at scale, across health systems, in a way that satisfies both EHDS requirements and pharma’s evidentiary standards. European organisations with credible governance frameworks and EHDS-ready infrastructure are the ones entering those conversations. The ones without are still pitching tools into a room that has moved on.
HLTH Europe 2026 confirmed the signals the programme had telegraphed. What Amsterdam added was a sense of operational urgency that the programme alone could not convey. The sector has moved from building frameworks to testing them in production. The failures are more public. The organisations doing serious work are identifiable. The distance between them and the rest is growing.
That is the kind of progress that is actually useful to track.
