Lord Arbuthnot of Edrom (Con)

My Lords, I declare my interest as a member of the Horizon Compensation Advisory Board. When, on 24 April this year, the noble Baroness, Lady Kidron, proposed an amendment to remove the presumption about the reliability of computer evidence, the noble Baroness who is now the Minister added her name to it—oh the perils of moving from opposition to government. My noble friend Lord Camrose—the Minister at the time—in a sympathetic speech, resisted that amendment on the basis, first, that there were shocking failures of professional duty in the Post Office case. This was quite true, but they were facilitated by the existence of the presumption. His second reason was that taking us back to the law of 1999, as the noble Baroness, Lady Kidron, eloquently set out just now, would risk undermining prosecutions because we would need to get certificates of accuracy in cases such as breathalysers and those involving emails. There may have been something in that, so the noble Baroness has proposed an amendment that is designed to get round that second point. I suspect that the Minister will resist this amendment too, but for reasons that I hope she will set out clearly, because we may then decide to move a different amendment on Report. We are making all the running on this—or at least the noble Baroness, Lady Kidron, is, with my full support and, I know, that of the noble Lord, Lord Clement-Jones. I take a moment out of this Committee to pay tribute to their work ethic in this Committee, which has been quite phenomenal. The Government do not seem to have the issue quite as close to the top of their priorities as we suggest. Without repeating all that I said on 24 April, I will summarise it as follows. Paul Marshall, the barrister, has pointed out that computer evidence is hearsay, with all the limitations that that implies. Modern computer programs are too large to be exhaustively tested. If computer programs are inherently unreliable, it is wrong to have a presumption that they are reliable. That issue will grow with the growth of artificial intelligence. The presumption that computer evidence is reliable leads either to such things as we saw occur in the Post Office scandal, with the Post Office essentially taunting the sub-postmasters, saying, “If you can’t show us what is wrong with the computer evidence, we don’t have to show you that evidence”—a shocking case of Catch-22; or to lawyers and courts voluntarily abandoning the presumption and denigrating all computer evidence, whether or not it deserves to be denigrated. That might lead, for example, to some defendants being acquitted when the evidence would require that they be convicted. We are trying to help the Government find a way through a problem that they recognise and assert exists. Will they please give us some help in return? This is both serious and urgent. Just saying that it is very difficult does not begin the process of putting it right.