Stop Hate Speech: identifying online hate speech using algorithms and strengthening public discourse

The Stop Hate Speech project was launched in 2018 by alliance F, the umbrella organisation of Swiss women's organisations (phase 1). It combines civil society engagement with technical approaches to develop and implement effective means against online hate speech. Since 2020, the Immigration Policy Lab at ETH Zurich and the Digital Democracy Lab at the University of Zurich have been providing the project with technological and scientific support (phase 2). The product is currently the most accurate algorithm in the German-speaking world for the automated detection of hate speech online. This enables a well-founded, quantitative assessment of the extent of the problem, which provides a basis for developing scalable countermeasures. Initial field experiments show positive effects of certain forms of ‘counter-speech’ - the targeted response to hate speech on the internet by people from civil society. In phase 3, which is now beginning, these approaches will be further deepened, and the project will deal even more holistically with strengthening public discourse online: in addition to combating hate speech, approaches will also be developed to promote particularly valuable online contributions.

• Phase 1 (Pilot; Q2 2018 - Q4 2020): completed
• Phase 2: (Stop Hate Speech; Q4 2020 - Q4 2022; funded by innosuisse): completed
• Phase 3: (Improving Public Discourse: Q4 2022 - Q4 2024): initialized

• alliance F
• Immigration Policy Lab, ETH Zurich
• Digital Democracy Lab, University of Zurich