Skip to main content

Trustworthy Artificial Intelligence

This area addresses the development of artificial intelligence systems that operate in a reliable, fair, and transparent manner, taking into account their social and environmental impact. Research in this domain focuses on the formal and methodological foundations required to evaluate and constrain the behavior of AI systems in specific contexts of use. It combines basic and applied research and supports the responsible deployment of AI in regulated and socially sensitive environments.

 

The main sublines include:

Algorithmic fairness

Study of methods for analyzing and mitigating bias in data and models, together with the formulation, comparison, and evaluation of fairness criteria in automated decision-making systems.

Computational and energy efficiency of algorithms.

Research on techniques that reduce the computational and energy cost of training and inference, with explicit consideration of trade-offs between efficiency, accuracy, and model complexity.

AI for sustainability and the green transition

Application of artificial intelligence methods to problems related to efficient resource management and sustainability, alongside the analysis of the environmental impact associated with the development and deployment of AI systems.

Casos de éxito