HealthAI maps EU AI Act gaps in health implementation

By AI, Created 10:18 AM UTC, May 19, 2026, /AGP/ – HealthAI released a report in Geneva on May 19, 2026, warning that the EU AI Act’s compliance deadlines are moving faster than the infrastructure needed to enforce them in health. The analysis focuses on Germany, France, Italy and Spain, and offers recommendations aimed at turning regulation into workable clinical adoption.

Why it matters: - The EU AI Act will shape how AI tools reach patients, hospitals and health systems across Europe. - HealthAI warns that compliance requirements are arriving before the support systems needed to make the rules work in practice. - The report argues that weak implementation could slow safe deployment, create enforcement uncertainty and limit AI’s competitive upside in health.

What happened: - HealthAI launched “Harnessing AI for Health and Economic Competitiveness: Translating the EU AI Act into Action” during the World Economic Forum’s Annual Health Roundtable on the margins of the 79th World Health Assembly in Geneva. - The report analyzes implementation challenges and opportunities for the European Commission and four member states: Germany, France, Italy and Spain. - HealthAI drew on 20 semi-structured stakeholder consultations, expert dialogues at the European Health Forum Gastein and policy analysis conducted from October 2025 through April 2026. - The report aims to improve institutional coordination and enforcement capacity for the AI Act’s rollout in health.

The details: - The report says implementation infrastructure is lagging behind compliance timelines. - Key support measures, including guidance on the AI Act’s relationship with medical device rules, the European Network of Expertise on AI Deployment in Healthcare and the full rollout of the European Health Data Space, are expected after, or at the same time as, high-risk provisions take effect. - The report identifies notified body capacity as a bottleneck. - As of October 2025, 51 notified bodies had been designated under the Medical Device Regulation. - AI Act designation requires additional expertise in AI technology and data science beyond traditional assessment skills. - National implementation is moving through different models. - Spain has created Europe’s first AI supervisory agency, AESIA, and launched the continent’s first regulatory sandbox. - Italy has enacted the EU’s first comprehensive national AI law. - Germany is pursuing dedicated implementation legislation. - France is embedding AI governance inside existing health regulation frameworks. - The report says no country has yet tested its dual supervision model in practice. - That model pairs horizontal AI authorities with sector-specific medical device regulators. - The report says market access, not just market authorization, may become the main barrier to adoption. - Germany and France have structured reimbursement pathways that can help scale AI tools in clinical settings. - Approved technologies in Spain and Italy often lack the budgetary and legal basis for adoption. - The report includes five strategic recommendations. - The recommendations call for joint exercises between AI authorities and medical device regulators before high-risk requirements take effect. - The recommendations urge harmonized reimbursement pathways across the EU so regulatory approval can lead to clinical use. - The recommendations ask the European Commission to clarify how the European Health Data Space, the AI Act and medical device regulation fit together. - The recommendations call for blueprints and roadmaps to guide stakeholders ahead of enforcement deadlines. - The recommendations propose structured peer learning across member states, facilitated by DG SANTE and the AI Office. - The recommendations also call for structured multistakeholder engagement at EU and national levels on oversight structures, conformity assessment and enforcement priorities.

Between the lines: - The report frames the EU’s challenge as one of execution, not ambition. - HealthAI’s analysis suggests Europe has built much of the policy architecture needed to lead in AI, but has not yet built enough operational capacity to support compliance. - The comparison across member states shows that national choices on supervision, reimbursement and legal structure may matter as much as the EU-wide rulebook. - The emphasis on reimbursement signals that regulatory approval alone may not move AI into routine care.

What’s next: - HealthAI wants countries to test dual supervision structures before the most demanding AI Act provisions take effect. - The European Commission is being urged to publish coordination tools before enforcement deadlines. - Member states are expected to use peer learning and multistakeholder engagement to close implementation gaps. - HealthAI presents the report as part of its role as an implementation partner for responsible AI governance in health.

The bottom line: - Europe’s AI-in-health question is no longer just what the rules say, but whether health systems can enforce, fund and operationalize them fast enough to matter.