Psychology · Politics · Data Science

Anton Gollwitzer

My research lies at the intersection of psychology, politics, and data science with a particular focus on destructive belief systems and social conflict. Recent work focuses on how polarization harms information integrity and how AI can be utilized to promote democratic life. My research is motivated by a commitment to protect and advance democratic institutions, civil liberties, and human rights.

I’ve published numerous peer-reviewed articles in leading journals, including Nature Human Behaviour and PNAS. My research has informed public policy discussions and received coverage in major international media outlets, including CNN, BBC, and The Times.

For collaborations, consulting, and speaking engagements, please contact: anton.gollwitzer@gmail.com

Curriculum Vitae CV of Failures

Areas of Research

All research

Deviancy Aversion

Does a low-level discomfort of broken patterns help explain our social behavior?

Learn more →

Social Psychological Skill

Can some people accurately judge social psychological phenomena without any background in psychology?

Learn more →

AI and Democracy

How does artificial intelligence threaten democratic life—and how can it be harnessed to strengthen it?

Learn more →

Motivation

How can we increase our motivation and succeed in our goals?

Learn more →

Fanaticism

What contexts and factors lead people to adopt fanaticism and extremism?

Learn more →

Misinformation

How can we reduce misinformation and combat false news?

Learn more →

Featured Research

All publications

Journal of Experimental Psychology: General2026

Toward a mechanistic understanding of false news sharing: Which interventions work best, for whom, and why

Gollwitzer, A., Tump, A. N., Martel, C., Deffner, D., Sultan, M., Kurvers, R., & Hertwig, R.

This study examines how to curb false news sharing, drawing on news-sharing decisions from over 1,100 U.S. participants. Comparing leading interventions head-to-head and using drift-diffusion modeling to uncover the decisions behind sharing, it finds that warning labels and media literacy tips substantially improve sharing quality, while social norm cues have a modest effect and accuracy prompts a very subtle one. These benefits hold broadly—even among at-risk groups like conservatives. Critically, each intervention works through a distinct cognitive pathway: warning labels shift people’s initial intentions away from sharing before they engage with content, whereas media literacy tips operate later, improving how news is processed and increasing caution before a decision. The findings clarify which interventions work best, for whom, and in which contexts.

Figure from “Toward a mechanistic understanding of false news sharing: Which interventions work best, for whom, and why”

PsyArXiv (preprint)2026

Misplaced Certainty in Elite Political Rhetoric Predicts Support for Political Violence

Gollwitzer*, A., Jeong*, Y., Von Damaros, M., Wilhelmsen, L. L., Pillai, S., Steinhauer, J., & Oettingen, G.

Five preregistered studies—spanning historical analyses, large-scale social-media data, and experiments—ask whether leaders asserting evidence-defying claims as certain fact fuels support for political violence. Misplaced certainty in elite rhetoric tracked later increases in violent political events, spread further online, and experimentally raised support for political violence along with its precursors: affective polarization, moral outrage, and perceived threat.

Figure from “Misplaced Certainty in Elite Political Rhetoric Predicts Support for Political Violence”

Personality and Social Psychology Review2026

Intelligent Systems, Vulnerable Minds: A Framework for Radicalization to Violence in the Age of AI

Kunst, J. R., Obaidi, M., Gollwitzer, A., Brandtzæg, P. B., Hinrichs, Y., Saini, N., & Schroeder, D. T.

This review outlines a four-stage framework for how AI reshapes the path to violent extremism. Engagement-optimizing algorithms and generative AI together build an “architecture of radicalization” unfolding through Exposure (recommender systems surface extreme content), Reinforcement (filter bubbles and bot-driven synthetic consensus cement beliefs), Group Integration (AI companions and bot swarms forge belonging and identity fusion), and finally Violent Extremist Action—with individual vulnerabilities like loneliness and need for closure moderating who is most at risk. We argue generative AI introduces novel dynamics, including “synthetic sociality” that decouples validation from human contact and a self-steered loop in which users unwittingly train the system to radicalize themselves.

Figure from “Intelligent Systems, Vulnerable Minds: A Framework for Radicalization to Violence in the Age of AI”

Proceedings of the National Academy of Sciences2024

Susceptibility to Online Misinformation: A Systematic Meta-Analysis of Demographic and Psychological Factors

Sultan, M., Tump, A. N., Ehmann, N., Lorenz-Spreen, P., Hertwig, R., Gollwitzer, A., & Kurvers, R.

This meta-analysis investigates susceptibility to misinformation, analyzing data from over 250,000 news judgments by U.S. participants. The study examines who can accurately identify true and false news, revealing that older adults, Democrats, and individuals with strong analytical thinking skills tend to have better accuracy. However, biases emerge: people more readily believe news that aligns with their political views or feels familiar. Republicans benefit more from having source information displayed, while overall, stronger critical thinking skills correlate with caution toward misinformation. The findings highlight the need for interventions that encourage unbiased, critical evaluation across demographics to counter misinformation.

Figure from “Susceptibility to Online Misinformation: A Systematic Meta-Analysis of Demographic and Psychological Factors”

Nature Human Behaviour2020

Partisan differences in physical distancing are linked to health outcomes during the COVID-19 pandemic

Gollwitzer, A., Martel, C., Brady, W. J., Pärnamets, P., Freedman, I. G., Knowles, E. D., & Van Bavel, J. J.

COVID-19 response in the US is deeply divided along partisan lines. Using data from 15 million smartphones across the United States, we found that counties that supported Trump showed 14% less physical distancing, and the divide grew over time. Consumption of conservative media correlated with reduced distancing, and these partisan differences led to higher infection and fatality rates in pro-Trump counties.

Figure from “Partisan differences in physical distancing are linked to health outcomes during the COVID-19 pandemic”

Journal of Experimental Psychology: General2020

Pattern deviancy aversion predicts prejudice via a dislike of statistical minorities

Gollwitzer, A., Marshall, J., & Bargh, J. A.

Research has shown a connection between people’s dislike of pattern deviations and their prejudice, but the underlying processes and causality remain unclear. We suggest that aversion to pattern deviations may contribute to prejudice by intensifying people’s dislike for statistical minorities. In nine studies (N=1,821), we found support for this idea, as aversion to pattern deviations related to disliking novel statistical minorities, which in turn predicted prejudice against racial and stigmatized minorities. These findings indicate that people’s aversion to pattern deviations can drive prejudice, particularly toward statistical minorities.

Figure from “Pattern deviancy aversion predicts prejudice via a dislike of statistical minorities”

Nature Human Behaviour2017

Relating pattern deviancy aversion to stigma and prejudice

Gollwitzer, A., Marshall, J., Wang, Y., & Bargh, J. A.

Research conducted across six studies indicates that people’s strong aversion toward non-social pattern deviancy, such as a row of triangles with one out of line, predicts their dislike of social deviancy. This relationship remains consistent across various measures and cultural contexts, displaying a moderate effect size. The findings suggest that aversion to pattern deviancy plays a significant role in shaping stigmatization and prejudice.