"This report examines political communication and media trust in the age of generative artificial intelligence systems (AI). Firstly, it provides a brief explainer of generative AI tools and techniques, looking separately at systems that generate text and those that generate or manipulate images, videos and audio. By reference to real-world examples, the paper then surveys the ways in which generative AI systems have recently been used by political actors, distinguishing between three different use-cases: political campaigning, entertainment and disinformation campaigns. Building on this empirical analysis, the paper distils important insights for policymakers, which highlight the need to: refrain from falsely labelling content as AI-generated to avoid overstating the technical capabilities and persuasive power of those spreading disinformation; acknowledge the multimodality of threats posed by generative AI, in particular voice-generation; delimit fair-use cases of generative AI for political campaigning, given these technologies are already widely used for legitimate political communication purposes; raise awareness of how seemingly non-political uses of generative AI can be exploited for politics, in particular the creation of non-consensual intimate content. This is followed by an evaluation of emerging technical and policy solutions, namely the detection and labelling of deepfakes as well as the development of systems to certify content authenticity and provenance. The section concludes with a discussion of the emerging legal landscape, including the European Union’s AI Act." (Executive summary)