Together with Maria Pawelec (IZEW, University of Tübingen), Mateusz published another peer-reviewed article titled “Are Softfakes Really ‘Soft’? Semantic, Ethical, and Legal Perspectives on Deepfakes and AI-generated Visualizations in Political Campaigning” in the International Journal of Politics, Culture, and Society.

Though one might not be familiar with that, “softfakes” is a label for deepfakes and AI-generated visuals that parties use to promote their own candidates. They are mainly employed for multilingual outreach, endorsements by deceased people, idealised avatars or chatbots that talk directly to voters. The authors argue the label is misleading as it leads to semantic softening of the dangerous phenomenon. Such uses can still manipulate and deceive voters, fragment political debate and raise concerns about human dignity. Calling them “soft” risks public indifference and might slow down regulation. For that reason, the authors call for updated political advertising rules, codes of conduct and stronger media literacy.

We are very happy to share this article as open access: https://link.springer.com/article/10.1007/s10767-026-09579-z