AI Project to Model Human Behavior and Epidemic Spread

Consortium Secures Funding for Artificial Intelligence Epidemic Models

A research consortium led by the University of Zaragoza has secured selection in the 2026 Health Research Call by the “la Caixa” Foundation to develop artificial intelligence models that replicate human behavior during infectious disease outbreaks. The AIBEM project received €970,139 over three years to build tools that simulate how trust in health authorities, misinformation, and socioeconomic inequalities influence the spread of disease.

Key Clinical Takeaways:
• The AIBEM project utilizes generative agent-based models to simulate adaptive human behavior during public health crises.
• Researchers from the University of Zaragoza, Universidad Carlos III de Madrid, and Universitat Pompeu Fabra will validate generative agents against behavioral experiments involving real human participants.
• Funding comes from a competitive 2026 biomedical research grant awarded by the “la Caixa” Foundation, which selected 32 proposals out of 800 submissions.

Three Spanish Institutions Divide Project Responsibilities

The three-year initiative divides responsibilities across three participating Spanish institutions. Universidad Carlos III de Madrid will conduct behavioral experiments with socioeconomically diverse participants using the IBSEN platform, drawing on experience from the European H2020 project of the same name. Anxo Sánchez of the university’s Interdisciplinary Group of Complex Systems stated that validating generative agents requires comparing their behavior against standard experiments and modeling epidemic scenarios. Meanwhile, the University of Zaragoza will develop the generative agents and integrate them into epidemiological models.

Yamir Moreno, director of the Institute for Biocomputation and Physics of Complex Systems and principal investigator for the project, noted that capturing human behavior during health crises remains an unresolved challenge. According to Moreno, equipping autonomous computational agents with reasoning capabilities allows researchers to transform epidemiological models into a faithful reflection of the human factor on a societal scale.

Integration of Social Determinants and Equity

Universitat Pompeu Fabra will lead the design and evaluation of interventions focusing on ecosocial determinants, health inequalities, and public policy. Joan Benach, director of the Health Inequalities Research Group, emphasized that epidemics are not purely biological phenomena because their spread is conditioned by social inequalities and living conditions. The project aims to ensure that technological models incorporate structural drivers that leave specific social groups more exposed.

To achieve this, age-based contact matrices will be expanded to include socioeconomic dimensions such as occupation, income, and neighborhood using Spanish census data and labor force surveys. This design ensures disparities in disease burden emerge directly from material conditions. Concurrently, arxiv.org published research on August 30, 2026, detailing a related framework named the Generative Adaptive Behavioral Layer for Epidemics, which adapts large language models to infer behavioral responses and couple them to mechanistic epidemic models. Applied to COVID-19 data in France, that framework outperformed traditional mobility-driven matrices in short-term forecasting.

AIBEM Project Ensures Open Access and Ethical Oversight

The AIBEM project incorporates an ethical and governance structure featuring bias audits, human oversight, and independent equity reviews. All models, datasets, and validation protocols will be published open-access following FAIR principles. Expected deliverables include behavioral agents validated against real experimental data, coupled epidemiological models quantifying how behavioral heterogeneity modulates transmission, and a decision-support toolkit for Spanish and European health authorities containing policy briefs, ethical guidelines, and simulation software.