ITU – Detecting Deepfakes and Generative AI, 2024

The 2024 ITU report outlines emerging standards for detecting deepfakes and ensuring multimedia authenticity through AI watermarking. This report presents outcomes from an ITU workshop that convened industry experts, standards organizations, and academics to address the growing risks associated with AI-generated content. Key points include the importance of international standards in AI watermarking, methods for verifying content authenticity, and the need for a robust policy framework. The report emphasizes collaborative approaches among governments, organizations, and tech providers to standardize methods for identifying AI-generated and manipulated media, supporting transparency and trust in digital content. Standardized approaches discussed include metadata, fingerprinting, and secure tracking tools that provide provenance information for multimedia.