Deepfakes and synthetic media have moved well beyond face swaps. This second edition offers an interdisciplinary examination of how synthetic media is created, detected, and governed, combining technical depth with global case studies. Organized into four parts covering foundations, advanced techniques, societal impact, and future directions, it has been updated throughout to reflect advances in multimodal generation, real-time detection, and adversarial robustness.
• Covers the full generative landscape, including voice cloning, cross-modal synthesis, and diffusion models, alongside established face-swap methods.
• Explains detection strategies from CNN-based classifiers to newer architectures, with attention to adversarial robustness, generalization to unseen generators, and explainability tools.
• Examines provenance and authentication approaches, including watermarking and the C2PA Content Credentials standard, together with consent frameworks and digital identity governance.
• Compares regulatory responses worldwide, including the EU AI Act's transparency obligations, US federal and state legislation and proposals, China's deep synthesis provisions, India's IT Rules on synthetic content, the African Union's Continental AI Strategy, and UNESCO- and UNDP-supported initiatives in the Middle East and Africa.
• Highlights constructive applications of synthetic media and offers a roadmap for future-proofing both technology and regulation.
This book is intended for graduate students, researchers, and professionals in artificial intelligence, computer vision, machine learning, and data science, as well as policymakers, legal scholars, and trust and safety practitioners working on digital media integrity.