AI-generated texts, images, and videos are becoming increasingly difficult to distinguish from authentic content. The transparency requirements of the European AI Act, applicable from 2 August, aim to address this challenge by requiring providers of AI systems to ensure that synthetic content is marked in a machine-readable format. Deepfakes are subject to additional transparency requirements indicating that image, audio, or video content has been artificially generated or manipulated.

In a new episode of the “Einstiegsdebatten” podcast by acatech – National Academy of Science and Engineering, Prof. Ahmad-Reza Sadeghi, discusses the technical feasibility and broader societal implications of these transparency requirements together with Prof. Armin Grunwald, Head of the Office of Technology Assessment at the German Bundestag (TAB) and Professor of Philosophy of Technology at Karlsruhe Institute of Technology (KIT).

The discussion explores whether reliable labeling of AI-generated content can be implemented technically, how such mechanisms can contribute to trust and authenticity online, and whether transparency requirements can help counter the growing difficulty of distinguishing synthetic from genuine digital content.

The episode is part of acatech’s podcast series on technologies shaping our future. The pilot season is accompanied by Plattform Lernende Systeme – Germany’s AI Platform as a cooperation partner.

Listen to the podcast:

https://lnkd.in/djxTMbFi