<![CDATA[Digital evidence, ranging from smartphone data and cloud-stored communications to surveillance footage, IoT sensor logs, and blockchain transaction records, has become central to modern criminal prosecutions, yet the legal frameworks that govern its use in court have struggled to keep pace with the technology that produces it. This article examines the persistent legal challenges surrounding the authenticity, reliability, and admissibility of digital evidence in criminal proceedings across common-law and civil-law jurisdictions. Employing a normative-doctrinal legal research method supported by a systematic review of contemporary scholarship, case law, and international instruments published mainly between 2017 and 2026, the study identifies three interlocking problems: the absence of harmonized statutory definitions and chain-of-custody standards for digital evidence, the epistemic difficulty of validating the reliability of forensic tools and artificial-intelligence-assisted analysis, and the emerging threat that synthetic media, commonly called deepfakes, pose to the long-standing presumption of authenticity attached to audio-visual records. The findings show that jurisdictions differ markedly in how they allocate the burden of authentication among investigators, forensic experts, and trial judges, and that inconsistent validation protocols for forensic software create a genuine risk of wrongful conviction or wrongful acquittal. The article further demonstrates that cross-border and cloud-based evidence collection exposes gaps in mutual legal assistance mechanisms that traditional procedural codes never anticipated. It argues that establishing binding minimum standards for digital forensic validation, strengthening the judicial gatekeeping function at the admissibility stage, and adopting technology-neutral yet procedure-specific rules of evidence are necessary steps to preserve the fairness, reliability, and legitimacy of criminal trials in an increasingly digitised world.]]>
