The effectiveness of the precautionary principle to protect the environment In achieving sustainable development

Abstract

This paper examines the role of Artificial Intelligence (AI) in digital criminal evidence analysis across investigation and trial stages, arguing that the growing complexity of cyber-related cases has made traditional methods insufficient in many situations. It highlights how AI can improve the speed and accuracy of criminal procedures by processing large-scale data, detecting patterns, supporting facial/image recognition, enabling predictive policing approaches, and extracting meaningful indicators from large volumes of communications and digital documents. The study also addresses AI applications in audio-visual evidence, such as voiceprint identification and smart surveillance analytics, while stressing the need to balance public security with defendants’ rights and privacy to ensure fairness. At the same time, it warns of serious risks including bias, inaccurate outputs, and manipulation of digital evidence—particularly through deepfake fabrication that may falsely attribute confessions or actions to innocent persons. The paper concludes with practical recommendations, including building specialized technical capacity, updating criminal procedure frameworks to accommodate modern tools, and limiting AI use whenever it