ADVANCING FORENSIC APPLICATIONS: THE ROLE OF ARTIFICIAL INTELLIGENCE IN DETECTING AND MITIGATING IMAGE AND VIDEO MANIPULATION

Abstract

Electronic information is threatened when forensic examinations become challenging due to modern digital advancements.  Through Deepfakes media manipulation forms the basis for people who either attack others or spread false information while influencing political opinions.  Traditional forensic testing methods show reduced capability in identifying picture alteration compared to more advanced techniques while operating at slow identification speeds.  Deep learning-based artificial intelligence serves as a strong tool for handling modern-day technical problems.  Our research assesses and develops artificial intelligence systems dedicated to detecting and reducing digital image and video manipulation for investigative purposes.  We execute data management followed by model development and training and final assessment of results throughout our program.  The developed AI models undergo verification using precision and accuracy together with recall testing as well as F1 score and ROC-AUC measurements.  Our assessment framework integrates the use of GANs with CNNs while also deploying CNN-RNN hybrid architectures.  AI-based techniques surpass traditional forensic methods to detect different types of manipulation with superior effectiveness.  Real-world application demonstrates these models function effectively while keeping adversaries from influencing them.  The research evaluates the cultural and ethical effects of artificial intelligence in forensic science while developing recommended methods of safe utilization.  The research uses artificial intelligence strengths but addresses its weaknesses to develop secure systems against digital media tampering.