RESEARCH PAPER ON ADMISSIBILITY OF AI-GENERATED EVIDENCE: A COMPARATIVE STUDY OF U.S., EU, AND U.K. LEGAL FRAMEWORKS

The recent adoption of artificial intelligence (AI) into investigative, forensic, and adjudicative procedures has put intricate issues on whether the evidence that is generated by AI should be admissible in a court of law. The paper at hand compares the approaches to evidence presentation, used in the United States, the European Union, and the United Kingdom when addressing the issue of AI-generated evidence, including the algorithms of risk evaluation, scan outputs, and automated data analyses. With the emphasis on background evidentiary principles, such as relevance, reliability, authenticity and fairness, the paper looks at how traditional doctrines are being modified (or stretched) with regard to opaque algorithm models, probabilistic reasoning and machine learning models that adapt as time passes. The bias of scientific validity and judicial gatekeeping in established standards of admissibility criterion is the focus of analysis in the United States with highlighting of tensions between innovation and due process in the American system. The close focus on essential rights, information protection, and transparency of the European Union is considered especially in the conditions of the new rules being developed to control high-risk AI systems. The structure of the United Kingdom is examined in terms of judicial discretion, common law latitude, and the emerging guidance on expert as well as digital evidence. The paper finds convergent issues, including the explainability, bias, and accountability, and divergent regulation through legislation or judicial ruling by comparing these jurisdictions. It states that although no jurisdiction has so far evolved a strictly consistent approach, specific to AI-generated evidence, comparative experience creates insights into best practices. Finally, the paper can be added to the greater discussion of the ways in which legal systems may protect procedural fairness and community trust and utilize AI-driven evidentiary tools in the present day litigation in a responsible manner.