Department of Communication and Culture

Proceedings of the National Academy of Sciences of the United States of America (PNAS) 123(13) Doi: https://doi.org/10.1073/pnas.2535665123

Evolutionary theory and historical evidence suggest humans possess distinct psychological tendencies for defensive and offensive violence, which have insufficiently been considered in research. In a large-scale preregistered study across 58 countries ( N = 18,128), we demonstrate that violent extremist intentions manifest along two distinct psychological phenomena: defensive extremism, motivated by protecting one’s group from (perceived) threats, and offensive extremism, driven by establishing group dominance. We show that these dimensions a) can be reliably differentiated across diverse cultural contexts, b) are distinctively associated with psychological dispositions, and c) systematically differentiate countries varying in macrolevel sociopolitical functioning and violence. Across nations, a two-factorial structure was observed that was invariant at the scalar level. Defensive extremist intentions were consistently higher than offensive extremism in 56 out of 58 countries, suggesting greater moral acceptance of protective violence. While psychopathy was positively related to both types of violent extremist intentions, those high in Machiavellianism and narcissism demonstrated particularly higher levels of defensive extremist intentions. By contrast, those scoring high on religious fundamentalism and social dominance orientation demonstrated particularly higher levels of offensive extremist intentions. Unexpectedly, liberal political group identification was associated with higher offensive but lower defensive extremist intentions. Crucially, offensive (but not defensive) intentions were associated with macrolevel societal dysfunction, including political terror and internal conflict. These findings establish that defensive and offensive violent extremist intentions represent two conceptually different forms of extremism across a large and diverse range of countries, with consequences for research and practice.

ACM Transactions on Social Computing (TSC) Just Accepted p. 1-27 Doi: https://doi.org/10.1145/3804460

Research has consistently shown that websites’ privacy policies fail to inform users, prompting to explore more condensed, visual solutions such as icons and privacy labels. However, we miss efficient and scalable systems to generate such privacy labels and conclusive evidence on their impact on people’s privacy protection behavior. This paper fills these gaps: we present the Visual Privacy browser extension that automatically computes a privacy rating and visualizes it in a privacy label. We studied the impact of those labels on privacy protection behavior in an online experiment ( N = 825). Our results show that the privacy labels shown by Visual Privacy significantly impact a website’s perceived risks and benefits, and that they have an indirect effect on privacy protection behavior. Study participants showed high awareness and acceptance of privacy labels, with 79% of participants recalling the rating shown in a label and over 75% reporting that they would like them to become mandatory.

Personality and Social Psychology Review Doi: https://doi.org/10.1177/10888683261430089

Academic Abstract Advances in AI require a revision of the psychological and socio-technical dynamics by which individuals are radicalized to embrace violent extremism. This review synthesizes process models of radicalization with research on social and personality risk factors, AI, and psychological mechanisms to propose a four-stage framework mapping the AI architecture of radicalization: (1) Exposure , where recommender systems and virality features create initial attraction to extreme content; (2) Reinforcement , where filter bubbles and group recommendations leverage biases to strengthen extremist beliefs and create echo chambers; (3) Group Integration , where ideologically homogenous clusters, AI bot swarms and companions foster group belonging and readiness for action; cumulatively resulting in (4) Violent Extremist Action . We examine how established social, cognitive, personality, and contextual vulnerability factors heighten psychological risk in the AI-driven radicalization process, as well as the emerging role of generative AI. We conclude by outlining a stage-based framework for governance and future research. Public Abstract AI-driven algorithms designed to maximize engagement on social media, compounded by generative AI, can unintentionally set the stage for radicalization. It begins with Exposure , where algorithms push users toward extreme content because it captures attention. Next, during Reinforcement , algorithms feed users personalized content while AI swarms can create a synthetic consensus that reinforces emerging biases, normalizes extremity, and insulates users from alternative views. Third, during Group Integration , individuals are absorbed into extremist networks, reinforced by human peers, AI companions, and bot swarms that validate radical beliefs and deepen identity ties. By exploiting psychological needs for belonging and certainty, this stage becomes particularly pernicious, potentially opening the door for violence. We propose policy measures that can reduce radicalization at each stage.

Scandinavian Journal of Educational Research p. 35-35 Doi: https://doi.org/10.1080/00313831.2026.2642596

Early numeracy skills are falling, leading to quests for new instructional approaches. Blended learning has potential to improve motivation and achievement in mathematics. We investigated the effects on pupil achievement and motivation of using Dragonbox School in 2019-2020, which aims for a holistic early numeracy approach with teacher training, books, physical concretes, and digital components. We used a quasi-experimental pre–post study across one school year, with an overall sample size (Ngrade 1 = 368 grade 1 and Ngrade 2 = 538 pupils), matching on prior data, and using structural equation modelling. Despite teacher satisfaction, we could not find convincing positive effects of the intervention, but found a negative development for motivation in grade 2. The findings imply that this blended learning approach is not a magic bullet, even with satisfied users. We discuss the lack of adaptive features and variable teacher competences as foundation for using the novel approach effectively.

International Journal of Strategic Communication p. 21-21 Doi: https://doi.org/10.1080/1553118x.2026.2617933