Dr. Manuel Tonneau

Areas of Research

  • Societal Impacts of AI

  • Evaluation of AI Systems

  • Online Harms and Content Moderation

Methods

  • Natural Language Processing (NLP)

  • Large Language Model Evaluation

  • Computational Social Science

  • Large-scale Data Analysis

  • Survey Research and Online Experiments

Biography 

A.T. (Adam) Kingsmith is a critical social scientist working at the intersections of AI governance, democratic accountability, and the political economy of digital technologies. He is a postdoctoral researcher in the Department of Digital Governance, Technology and Policy at the Hasso Plattner Institute, where his work focuses on algorithmic governance, digital sovereignty, and the democratic implications of generative AI. Alongside this, he brings an active research focus on gamification and platform engagement, examining how game mechanics are deployed within digital systems to shape behaviour, manufacture consent, and modulate political subjectivity. Prior to joining HPI, he held research and teaching appointments at Humber Polytechnic, OCAD University, and Toronto Metropolitan University, and served as Technical Advisor on AI Ethics and Policy at the Digital Governance Standards Institute of Canada.

His work has appeared in Big Data & Society, the Journal of Psychoanalytic Studies, Global Political Economy, Theory, Culture and Society, and South Atlantic Quarterly, among others, and has been featured in public outlets including The Conversation, CBC News, Mad in America, Public Seminar, and the Los Angeles Review of Books. His forthcoming book, Anxiety as a Weapon, is due out from Athabasca and University of Chicago Press in Fall 2026.

Recent Publications

First-author work. Full list on Google Scholar.

  • Tonneau, M., Liu, D., Malhotra, N., Hale, S. A., Fraiberger, S. P., Orozco-Olvera, V., & Röttger, P. (2025). HateDay: Insights from a Global Hate Speech Dataset Representative of a Day on Twitter. ACL 2025. Outstanding Paper Award. 10.18653/v1/2025.acl-long.115

  • Tonneau, M., Sehgal, N. K. R., Malhotra, N., Kazemi, S., Orozco-Olvera, V., Muñoz Boudet, A. M., Subramanian, L., Fraiberger, S. P., Guntuku, S. C., & Hofmann, V. (2026). Different Demographic Cues Yield Inconsistent Conclusions About LLM Personalization and Bias. EMNLP 2026. https://doi.org/10.48550/arXiv.2601.18486

  • Tonneau, M., Liu, D., McGrady, R., Zheng, K., Schroeder, R., Zuckerman, E., & Hale, S. A. (2026). Language Disparities in Moderation Workforce Allocation by Social Media Platforms. FAccT 2026. https://doi.org/10.1145/3805689.3806484

  • Tonneau, M., Quinta de Castro, P. V., Lasri, K., Farouq, I., Subramanian, L., Orozco-Olvera, V., & Fraiberger, S. (2024). NaijaHate: Evaluating Hate Speech Detection on Nigerian Twitter Using Representative Data. ACL 2024.

  • Tonneau, M., Liu, D., Fraiberger, S., Schroeder, R., Hale, S. A., & Röttger, P. (2024). From Languages to Geographies: Towards Evaluating Cultural Bias in Hate Speech Datasets. WOAH 2024. 10.18653/v1/2024.woah-1.23

Education

  • York University, Toronto, Canada. Ph.D. in Political Science (International Relations and

    Political Economy) — 2023

  • McMaster University, Hamilton, Canada. M.A. in International Relations — 2012

  • Simon Fraser University, Vancouver, Canada. B.A. in Political Science — 2011 PhD in Social Data Science, Oxford Internet Institute, University of Oxford, 2022–2026. Thesis: “Examining Disparities in Online Hate Speech Moderation across the Globe”, supervised by Ralph Schroeder and Scott Hale.

  • Engineering degree in Statistics and Economics (equivalent to MSc), ENSAE Paris, 2016–2019, including an MSc in Economics at Humboldt-Universität zu Berlin as part of a double degree programme.

  • BSc in Economics and Applied Mathematics, ENSAE Paris, 2015–2016.

  • Classes Préparatoires aux Grandes Écoles in social sciences and humanities, Lille, 2012–2015.

Experience 

  • Postdoctoral Researcher, Business and Society Cluster, Hasso Plattner Institute, University of Potsdam, since September 2026.

  • Consultant, Development Impact group, The World Bank, 2020–2026. Research on online harms and AI evaluation in Global Majority contexts.

  • Associated Researcher, Weizenbaum Institute, Berlin, since October 2024 (Platform Algorithms and Digital Propaganda group).

  • Research Assistant, University of Exeter Business School, 2024. Study of moderation rules on the Fediverse.

  • Research Engineer, Computational Social Science team, Centre Marc Bloch, Berlin, 2021–2022.

  • NLP Research Engineer, Creatext, Berlin, 2019–2020.

  • Supervision: MSc thesis supervision at the Oxford Internet Institute and research assistants at the Weizenbaum Institute.

  • Reviewing: ACL Rolling Review, ACL workshops (WOAH, NLP+CSS, SemEval), LREC, IC2S2, ACM Journal on Computing and Sustainable Societies.

Award and Recognition

  • Oxford Internet Institute – Dieter Schwarz Foundation Award, December 2025. One-year research award on “AI, Government and Policy” (£24,420 stipend and £4,000 research allowance).

  • Outstanding Paper Award, Association for Computational Linguistics, July 2025, for “HateDay”. One of 26 awards among 1,699 accepted papers, from 8,360 submissions.

  • Bronze Medal, Expedition Aya 2025, Cohere Labs, May 2025. Co-led a team of eight on political bias in large language models across languages, ranked third of more than 25 projects.

  • Visiting Fellowship, Weizenbaum Institute, October 2024 – February 2025 (€12,000), hosted by the Platform Algorithms and Digital Propaganda group.

  • Global Merit Award, Shirley Scholars Fund, Oxford Internet Institute, October 2022. Merit-based departmental scholarship covering doctoral tuition (£84,096).

Selected Presentations

  • “HateDay: Insights from a Global Hate Speech Dataset Representative of a Day on Twitter”. Oral presentation at ACL 2025.

  • “Assessing global hate speech moderation on Twitter”. Invited talks at the Meta Computational Social Science seminar, the Stanford Security Lunch Seminar, the Data Science Brown Bag Series at the Hertie School of Governance, the MilaNLP lab at Bocconi University, Royal Holloway and the Weizenbaum Institute. 2024–2025.

  • “Characterizing networks of hate on Nigerian Twitter”. Presented at IC2S2 2023 in Copenhagen, Denmark.

  • “Cultural bias in hate speech datasets”. Presented at WOAH at NAACL 2024 and at C3NLP at ACL 2024.