How should the courts make decisions about the reliability of computer-derived evidence? Rebutting the rebuttable presumption of reliability
The current doctrine is that material from a computer to be used in legal proceedings is presumed to be reliable unless there is evidence to rebut that presumption, at which point the person wanting to rely on that material has to find ways to persuade a court that it is reliable. It is usually for a defence team to make the rebuttal and they may not always find it easy to do so. The doctrine came under strong questioning in the UK Post Office Horizon cases where courts were obviously being misled by assertions from the Post Office and others as a result of which many sub-postmasters were punished in civil and criminal proceedings and some committed suicide.
But what can replace the “rebuttable presumption of reliability”? An older doctrine said that the person responsible for the computer had to provide a certificate that it was working normally. If there was no certificate the computer material was regarded as inadmissible. S69 Police and Criminal Evidence Act 1984. But that scheme failed and was abandoned on several practical grounds – how could you find a plausible certifier for a commercial system that might consist of several different elements and be very complex?
In January 2025 the UK Ministry of Justice issued a call for submissions on The use of evidence generated by software in criminal proceedings.
My response provides a history of the various arguments and reviews possible reforms. In the end I opt for an arrangement by which some-one intending to tender computer-derived evidence is required to complete a questionnaire issued under the Practice Directions. The questions help to establish the level of reliability. I consider this approach preferable to attempting statutory definitions against which a court will have to judge admissibility.