From AI to Admissible: Implications of Computer-Generated Evidence in Commonwealth Courts

The increasing reliance on digital and computer-generated evidence, including Artificial Intelligence (AI) and Automated Technology (AT), poses significant challenges for legal systems across Commonwealth jurisdictions. This paper examines the evolving landscape of computer and AI-generated evidence, focusing on its admissibility and reliability within the judiciary. Historically, frameworks like England and Wales’ Police and Criminal Evidence Act (PACE) have guided the admissibility of electronic evidence, but these frameworks may now be outdated given the rise of AI technologies that operate autonomously and with minimal human input. AI-generated evidence introduces complexities not traditionally seen with earlier forms of computer evidence. Issues of authenticity, bias, and transparency arise due to the opaqueness of AI algorithms and the potential for manipulating digital data. This paper explores these concerns, using case studies such as the Horizon Post Office scandal in the UK and AI-driven surveillance in India to illustrate the risks posed by uncritically accepting AI evidence. Furthermore, the paper compares the English and Welsh approach, which assumes digital evidence to be inherently reliable, with the Canadian system, which demands a higher standard of proof for electronic evidence under the Uniform Electronic Evidence Act (UEEA). Both frameworks, however, are found lacking in addressing the nuances of AI-generated evidence. The paper concludes by advocating for the reform of existing legal frameworks to better account for the challenges posed by AI technologies. By establishing more rigorous standards for the admissibility of AI-generated evidence, Commonwealth legal systems can better safeguard the integrity of judicial proceedings and ensure fairness in an increasingly digital age.