AUTHENTICITY AND CHAIN ​​OF CUSTODY OF DIGITAL EVIDENCE IN BRAZILIAN CRIMINAL PROCEEDINGS: RISKS OF MANIPULATION BY ARTIFICIAL INTELLIGENCE

Abstract

This article analyzes the authenticity and chain of custody of digital evidence in Brazilian criminal proceedings, focusing on the risks of manipulation arising from artificial intelligence (AI) technologies, particularly deepfakes. The increasing digitalization of human interactions poses new challenges for the production and assessment of criminal evidence, requiring robust mechanisms to preserve evidentiary integrity. The central question is: how can digital evidence maintain its integrity and reliability in the face of falsification risks amplified by AI? The general objective is to analyze the importance of the chain of custody in preserving the authenticity of digital evidence, in light of Law No. 13.964/2019. Specifically, the study aims to: understand the normative foundations of the digital chain of custody; analyze validity criteria for electronic evidence; evaluate the impact of AI on evidentiary credibility; identify control mechanisms such as blockchain and encryption; and reflect on future perspectives. This is a qualitative study based on bibliographic and documentary research, employing the deductive method. The results indicate that the statutory formalization of the chain of custody via the "Anti-Crime Package" was a necessary step, yet insufficient given the lack of technical standardization and the rapid development of synthetic manipulation tools. The study concludes by highlighting the urgent need for integrated legal-technological governance, the training of legal practitioners, and the adoption of forensic protocols aligned with the constitutional principles of due process, the adversarial system, and the right to a full defense.