Mediated society (MEDIS:ON)
The project Mediated Society (MEDIS:ON) responds in both scope and objectives to the dramatic technological transformations that have become embedded in everyday life in modern societies. MEDIS:ON focuses on investigating the societal impacts of artificial intelligence, viewing AI both as a tool that expands human capabilities and as a source of new challenges, including the deepening digital divide, the spread of conspiracy narratives and deepfakes, and insufficient development of AI literacy. The primary aim of MEDIS:ON is to advance and deepen our understanding of processes related to information handling at the individual, social, and societal levels. This is approached through the perspectives of selected social sciences and humanities disciplines — including psychology, political science, theology, and education — in close collaboration with computer science.
From classroom to the city: Sociometry for data-driven school network management (SchoolNet)
Funded by the Technology Agency of the Czech Republic (TAČR) under the SIGMA programme, the SchoolNet project aims to improve evidence-based decision-making in school safety and prevention. Addressing the lack of consolidated data on school social safety and climate, the project will develop and validate a secure and ethical framework for transferring anonymized sociometric data from schools to local authorities. It will identify comparable indicators across schools, establish standards for protecting minors' data, and examine how access to such information can improve preventive policies. A key outcome will be an enhanced version of the YoungLink software, enabling local authorities to monitor and compare aggregated data on classroom and school climate. The project will also produce methodological guidelines and user manuals.
THE SUBSTITUTION: How GenAI Reconfigures Human Agency and Learning (SUBSTITUTION)
Generative artificial intelligence is rapidly shifting from a cognitive support tool toward a system capable of substituting core components of human thought. This project develops the first mechanistic, multilevel account of how sustained interaction with GenAI reshapes human cognition, agency, and moral responsibility — a process we term epistemic substitution. Across three coordinated work packages and seven preregistered studies, the project traces a theoretical cascade from distorted confidence to behavioural dependence to diffuse moral accountability. Work Package I examines whether AI interaction inflates perceived relative to actual understanding, when individuals begin delegating cognitive work to AI, and how delegation affects physiological indicators of cognitive engagement. Work Package II investigates whether repeated delegation stabilises into self-reinforcing behavioural dependence through reward-based mechanisms. Work Package III examines whether epistemic delegation extends into the moral domain, diffusing personal responsibility in human–AI decision systems. The project combines large-scale randomised online experiments with laboratory studies using eye-tracking, EEG, and cardiovascular measurement, analysed within a Bayesian multilevel framework. All studies are preregistered and conducted in accordance with open science principles. Findings will directly inform educational policy, AI governance, and the design of responsible human–AI systems.