A research consortium led by the University de Zaragoza and the Universidad Carlos III de Madrid has secured €970,139 in the 2026 Health Research Call from the “la Caixa” Foundation to build artificial intelligence models that simulate human behavior during disease outbreaks.
Foundation Selects Project from Eight Hundred Proposals
The “la Caixa” Foundation selected the project from a field of 800 proposals, yielding a 4% success rate across 32 total awardees sharing €25.7 million. The University de Zaragoza leads the initiative, deploying the supercomputing infrastructure at the Institute for Biocomputation and Physics of Complex Systems alongside the Complex Networks and Systems Group.
Universidad Carlos III de Madrid contributes through its Interdisciplinary Group of Complex Systems, utilizing the IBSEN platform to run behavioral experiments with diverse human participants. Universitat Pompeu Fabra joins through the Health Inequalities Research Group and the JHU-UPF Public Policy Center, integrating ecosocial determinants into the predictive framework.
Generative Agent-Based Models Replace Static Parameters
Traditional epidemiological models treat human adherence to vaccination, masking, and mobility restrictions as fixed or exogenous values. The AIBEM project replaces this approach with generative agent-based models powered by large language models, similar to the technology underlying ChatGPT and Gemini.
Yamir Moreno, principal investigator and director at BIFI, stated that capturing human behavior has historically been an unresolved challenge. “The rise of generative AI changes the rules of the game in our favor: by equipping autonomous computational agents with reasoning capabilities and mutual influence, we can transform epidemiological models from mere mathematical projections into a faithful reflection of the human factor on a societal scale,” Moreno said.
Integrating Socioeconomic Inequities into Disease Transmission
The project expands standard age-based contact matrices to incorporate occupational, income, and neighborhood data sourced from Spanish census records, labor force surveys, and stratified epidemiological data. This design ensures that disease burden disparities emerge naturally from material conditions rather than serving as an afterthought.
Joan Benach of Universitat Pompeu Fabra noted that epidemics remain deeply connected to social inequalities and power dynamics. “Our challenge is to ensure that AI goes beyond better predicting epidemics to helping explain them and developing socially just public health responses—built with stakeholder participation and capable of reducing, rather than reproducing, existing inequalities,” Benach said.
Researchers Compare Simulated Outputs Against Real Human Experiments
Validating the generative agents requires comparing their simulated outputs against data collected from real human experiments. Anxo Sánchez of Universidad Carlos III de Madrid emphasized that designing these scenarios is crucial for the project’s success. “This capability is key, as validating generative agents requires not only comparing their behavior against standard experiments, but also modeling epidemic-related scenarios and designing experiments that faithfully replicate them,” Sánchez noted.
The project framework complies with the European AI Act, incorporating bias audits, human oversight, and independent equity reviews. All models, datasets, and validation protocols will be made available via open-access following FAIR principles, culminating in a decision-support toolkit for Spanish and European health authorities.
Frequently Asked Questions
- How much funding did the AIBEM project receive?
- The project was awarded €970,139 by the “la Caixa” Foundation over a three-year implementation period.
- What specific technology powers the behavioral agents?
- The project uses large language models, the same underlying technology as AI platforms like ChatGPT and Gemini, to configure simulated individuals with specific profiles, memories, and goals.
- Which institutions make up the research consortium?
- The consortium is led by the University de Zaragoza, with participation from Universidad Carlos III de Madrid and Universitat Pompeu Fabra.