Yamir Moreno, BIFI Director, Secures €970,139 for Pioneering Project Using AI to Predict Public Behaviour During Epidemics

The ”la Caixa” Foundation has selected AIBEM as one of 32 projects to receive funding under its 2026 Health Research Call, which attracted 800 applications. This marks the first time the University of Zaragoza has led a project selected under this call, a milestone for research in Aragon.

Yamir Moreno, Professor at the University of Zaragoza and Director of the University Institute for Biocomputation and Physics of Complex Systems (BIFI), has secured €970,139 to lead AIBEM (AI-Driven Behavioural Epidemic Models for Intervention Design), a project that will use artificial intelligence (AI) to predict how people behave during epidemics.

Moreno received the award at a ceremony held in Barcelona on 15 October, accompanied by Pilar Pina, Vice-Rector for Scientific Policy at the University of Zaragoza, and Anxo Sánchez of Carlos III University of Madrid (UC3M) and Joan Benach of Pompeu Fabra University (UPF), who lead the partner teams involved in the project.

The COVID-19 pandemic highlighted how compliance with public health measures varies between social groups and over time. However, conventional epidemiological models still treat human behaviour as a fixed parameter.

How AIBEM will work

The project will develop generative agent-based models (GABMs), in which each simulated individual is powered by a large language model, such as those used by ChatGPT or Gemini. Each agent will have its own profile, including occupation, income, neighbourhood, beliefs and trust in the authorities, as well as its own memory and objectives. The agents will interact with one another, make decisions and potentially develop social norms.

Their behaviour will be tested through experiments involving real people and incorporated into epidemiological models to examine how these differences affect disease transmission and the effectiveness of interventions. The project will ultimately deliver a set of open-access tools enabling Spanish and European public health authorities to assess interventions before implementing them.

Equity at the heart of the project

AIBEM will examine not only whether an intervention works on average, but also which groups benefit from it and which do not. The models will incorporate socioeconomic variables to assess how living conditions influence different groups’ exposure to disease and their ability to protect themselves.

The project will also establish an ethical and governance framework aligned with the European Union’s Artificial Intelligence Act (EU AI Act), including bias audits, human oversight and independent equity reviews. The models, data and validation protocols will be made openly available in line with the FAIR principles.

Three teams, three areas of work

AIBEM will run from November 2026 to October 2029. Carlos III University of Madrid (UC3M), through Anxo Sánchez’s team, will conduct behavioural experiments with participants from diverse socioeconomic backgrounds using the IBSEN platform. The University of Zaragoza, through the BIFI’s COSNET group, will develop the generative agents and integrate them into epidemiological models. Finally, Pompeu Fabra University (UPF), through Joan Benach’s team, will turn the findings into decision-support tools for public health authorities and assess how interventions affect different social groups.