AI incidents and hazards in the media: Evidence from German-language coverage

Media reports of artificial intelligence (AI) incidents and hazards can shape public perception and political debates. However, both the level and the thematic focus of media attention vary across regions.
This paper analyses live data from the OECD.AI Incidents and Hazards Monitor (AIM) to examine how the distribution of 15 thematic clusters in German-language media compares with global trends. The analysis identifies three themes that are overrepresented in German-language coverage – autonomous vehicles, intellectual property, and large language models (LLMs) – and two themes that are underrepresented – synthetic media, and cyberattacks and fraud. Coverage of the remaining ten thematic clusters is broadly aligned with global media patterns. Overall, the findings suggest that media representation of AI risks is shaped not only by the nature of AI incidents but also language-specific and regional contexts.


























