Artificial Intelligence is playing an increasingly important role in innovation in health research and healthcare, from prevention and early detection to diagnosis, treatment and long-term disease management. This report presents an in-depth analysis of how ERC-funded frontier research is advancing AI-based approaches across the health domain.
The report examines ERC projects that develop and apply AI models, clinical decision support systems and platforms, including machine learning and deep learning. These projects aim to enable earlier disease detection and to tailor risk prediction, diagnosis, prognosis, treatment and disease management to individual patients.
AI is also shown to support the integration of multi-omics, phenotypic and health data, and to play a growing role across the lifecycle of medicines, from drug discovery to clinical trials. The report follows two earlier F2P analyses that mapped 1 048 ERC projects involving AI and explored their relevance for EU policy, as well as ERC researchers’ views on the future of AI in science and the impact of generative AI.
ERC portfolio on AI in health
Focusing specifically on AI in health, the report analyses 238 ERC-funded projects under FP7, Horizon 2020 and Horizon Europe (up to the 2022 calls), representing a total of EUR 450 million in funding.
With 91 projects in Life Sciences, 101 projects in Physical Sciences and Engineering, and 38 projects in Social Science and Humanities, the ERC portfolio sheds light on the breadth of scientific domains contributing to AI in health.
Artificial intelligence (34%) and applied computer science (20%) are the most prominent disciplines, followed by cognitive neuroscience (16%) and neuroscience (14%), including neurology. Machine learning is the most prevalent topic in over 45% of projects, while computational modelling and simulation methods stand out as key methodological approaches in 60% of projects.
Policy context: AI and health in the EU
The report is published at a time of rapid developments in the European policy framework for AI. Since the launch of the European AI Strategy in 2018, EU leaders have placed AI and its applications on the political agenda, aiming to boost Europe’s competitiveness while safeguarding fundamental rights and democratic values.
The AI Act, a flagship initiative, classifies most AI-based software intended for medical purposes as ‘high-risk’, subject to strict requirements on risk management, high-quality data sets, transparency and meaningful human oversight. In parallel, the European Health Data Space (EHDS) aims to improve citizens’ access to and control over their electronic health data, while enabling secure, privacy-preserving reuse of data for public health, policy and research.
Complementing these initiatives, the EU’s Apply AI Strategy, adopted in October 2025, and the European Strategy for AI in Science, set out a joint blueprint for responsible AI deployment in strategic sectors, including healthcare, pharmaceuticals and biotechnology, and for harnessing AI to accelerate scientific discovery.
Against this backdrop, the ERC portfolio on AI in health offers timely evidence on how frontier research can inform regulation, support implementation and help ensure that AI in health is trustworthy, effective and aligned with European values.
Case studies and challenges
A deeper dive into 59 ERC projects on AI in health, carried out with support from the Horizon Results Booster and in collaboration with the European Health and Digital Executive Agency (HaDEA), provides further insight into results and impact. The analysis reviews publications, clinical studies, devices and prototypes, patents and wider social, technological, economic and policy effects.
Twenty case studies demonstrate how ERC grantees are deploying AI across six key application areas: disease detection and monitoring, drug discovery, forecasting of risk factors and outcomes, imaging, medical robots and personalised medicine. These examples highlight both scientific breakthroughs and innovation pathways, including spin-offs and additional funding.
In interviews, researchers point to major challenges for AI in health, including fragmented and inconsistent biomedical data, difficulties in accessing and integrating sensitive health data across borders, and the need for rigorous validation and transparent integration of AI into clinical workflows. Concerns about accountability, bias, explainability and skills gaps also emerge strongly, particularly for dynamic, self-learning systems used in decision support.
Support needs and future directions
Looking ahead, the report relays clear calls from ERC researchers for stable, long-term funding to sustain frontier science and support translation from early-stage discoveries to clinical applications. They highlight the importance of secure, high-performance computing and data infrastructures, access to large, high-quality longitudinal health datasets, and European sovereignty over health data and AI.
Researchers also advocate for AI-for-science hubs and other ecosystems that join up funding, infrastructure, regulation and talent, along with clearer, more agile EU regulatory frameworks and dedicated AI sandboxes to safely test innovative solutions in real-world settings.
Read the full report
