The European Parliament's Committee on Public Health (SANT) on 24 June 2026 debated the role of artificial intelligence in healthcare, revealing divergences on regulatory stringency, data use, and workforce substitution. Chair Adam Jarubas (EPP) opened the debate by stressing that AI deployment outpaced regulation and required stronger data infrastructure and clinician involvement. Guest speakers Maciej Malawski (Sano Centre) argued for large diverse datasets and sovereign European AI, Sarada Das (CPME) warned that digital policy was fragmented and insufficiently grounded in healthcare realities, and Andrea Luca (Galeazzi Hospital) presented digital innovation as a clinical shift to ease workforce shortages.
Tomislav Sokol (EPP) questioned whether AI standards were too restrictive, while Ignazio Roberto Marino (Greens/EFA) and Catarina Martins (The Left) warned against over-reliance on AI for diagnosis and workforce substitution. On data use, Gerald Hauser (PfE) insisted on patient consent and refusal rights, while Malawski defended anonymized data for research. Das highlighted unresolved EHDS issues on opt-outs and liability. On interoperability, Vytenis Povilas Andriukaitis (S&D) stressed the need for standardized European methods, and Malawski noted hospitals were not ready. On human contact, Hauser and Andriukaitis warned against dehumanization, while Das advocated augmented intelligence. On sovereignty, Nicolás González Casares (S&D) and Malawski pushed for European AI infrastructure, while Das cautioned against shortcuts. On workforce relief, Luca saw potential, but Martins and Hauser argued AI cannot solve shortages.
Consensus emerged on AI's importance, need for better data, patient safety, AI as support for clinicians, and stronger European capacity. Jarubas concluded that AI could assist but not replace professionals, and the committee would revisit the issue. The next SANT meeting is scheduled for 29 June.
stricter AI standards could protect patients but risk stifling innovation, while reliance on anonymized data boosts research but may conflict with patient consent rights. European AI infrastructure would reduce dependence on non-EU tech but require significant investment. Workforce relief from AI could ease shortages but may dehumanize care if overused. Stakeholders most impacted include EU healthcare providers (facing implementation costs), patients (balancing safety and access), tech companies (regulatory burden vs. market opportunities), and EU member states (funding and interoperability challenges).