Book Title: Computational Criminology: AI Applications in Forensic Science and Criminal Justice
Editors: Dr. Xavier Louis, Dr. Surbhi Girdhar, Ms. Aswathi Chandran Nair, Mr. Ravi Kumar, and Ms. Nandini Katare
Chapter: 20
DOI: https://doi.org/10.59646/704/20
Author: Reshma M Ashok
Abstract
The emergence of generative adversarial network (GAN) and diffusion model-based synthetic media colloquially termed ‘deepfakes’ poses a fundamental challenge to the integrity of digital evidence in criminal proceedings. Video footage, audio recordings, and photographic images that have historically served as among the most compelling categories of forensic evidence are now susceptible to highly realistic synthetic fabrication, potentially enabling the creation of false alibis, fabricated victim statements, manufactured confessions, and fictitious crime documentation. This chapter examines the generative architectures that enable synthetic media creation including face-swapping GANs, neural voice cloning, and diffusion-based image synthesis surveys the computational detection methods developed to distinguish authentic from synthetic media, evaluates the forensic validation status of these detectors, and analyses the evidentiary, procedural, and systemic implications of deepfake proliferation for the reliability of digital evidence in criminal and civil proceedings. The chapter argues that deepfakes compel a fundamental reassessment of the evidentiary status of digital media and demand new authenticity verification standards, forensic tool development, and judicial protocols.