The Law Commission and the presumption that computer evidence is reliable

Since I started taking an active interest in the Post Office scandal, some three years ago, I have been impressed by the quality of writing and clarity of thought from lawyers. However, there are two major, linked exceptions.

The first exception was the comments from old school judges, some of whom almost seemed to take pride in the ignorant waffle about computers that they airily spouted from the bench. See Paul Marshall’s “The harm that judges do” for some excruciating examples.

The second exception concerned the presumption in England and Wales that computer evidence is reliable. The old school nonsense and this presumption of reliability came together in the work of the Law Commission in the 1990s. The presumption is phrased in these terms.

“In the absence of evidence to the contrary, the courts will presume that mechanical instruments were in order at the material time.”

That quote was a recommendation from the Law Commission’s 1995 consultation paper “Evidence in Criminal Proceedings: Hearsay and Related Topics” (see paragraph 14.28). It might seem that this has little to do with complex software systems, but the presumption applies to computer evidence. The legal establishment assumes, or has to pretend, that massively complex, distributed systems are basically “mechanical instruments”.

In 1997 the Law Commission followed up its consultation paper with a final report “Evidence in Criminal Proceedings: Hearsay and Related Topics”. This confirmed the initial recommendation. Parliament waved through the recommendation with little serious scrutiny and the presumption came into effect in 2000.

In 2020 I wrote an article for the Digital Evidence and Electronic Signature Law Review in which I explained why I thought the presumption of computer reliability was nonsensical and dangerous, “The Post Office Horizon IT scandal and the presumption of the dependability of computer evidence”.

I read the Law Commission’s consultation paper and final report while researching my article. I was surprised by the weird arguments presented in favour of the repeal of section 69 of the Police and Criminal Evidence Act 1984 and the resulting presumption of computer reliability. Perhaps weird is too kind. The quality of the case presented for repeal was wretchedly poor.

Rather than citing technical experts in support of repealing section 69, the Law Commission repeatedly quoted vague, arm-waving, un-evidenced comments by judges who offered no insight into anything beyond their own technical ignorance. Where the Commission did cite experts the quotes did not offer any support: quite the reverse – they undermined the case for repeal.

I mentioned this to Stephen Mason, editor of the Digital Evidence and Electronic Signature Law Review, and speculated about the basis for the Law Commission’s recommendation. I thought it would be interesting to follow the trail of evidence and assess the strength of Commission’s arguments. Stephen thought this would make an interesting article and commissioned me to write it.

What I found astonished me. The Law Commission’s work would have been inadequate for an undergraduate essay. An internal audit report of such abysmal quality would have seriously damaged the credibility, and the career, of the culprit. No respectable internal audit department would have allowed it to be issued. This was my verdict in the article’s conclusion.

“The Law Commission had an inadequate understanding of a complex technical subject and provided no credible justification for the presumption. Its 1997 report was incompetent, with conclusions and recommendations only loosely connected to findings, and findings that had no basis in fact.

The report shows a consistent pattern of sources that were misunderstood or misrepresented, and a persistent failure to appreciate the significance of the distinction between admissibility of computer evidence and its reliability. The resulting change to the law has had devastating consequences for thousands of people. It is shocking that this was brought about by a report of such abysmal quality.”

I did not have to carry out any clever or sophisticated investigation. I did not submit any Freedom of Information requests. All of the sources I consulted were already publicly available. I had only to perform internet searches, buy access to two academic articles and make contact with two of the main sources on whom the Law Commission relied, and whom they misrepresented. Alistair Kelman and Stephen Castell both confirmed that my analysis was correct. They had argued for PACE section 69 to be replaced with regulation that would require the providers of computer evidence to demonstrate that the systems had been developed and managed responsibly. They were both opposed to a presumption that systems were reliable.

My draft article was reviewed by various IT and legal experts, including one who was familiar with the work of the third source, Professor Colin Tapper. Only then was my article published.

Did nobody want to challenge the Law Commission? Did the Commission assume it was above reproach and no-one would check its work? It still amazes me that such a shoddy and damaging piece of work escaped serious scrutiny for quarter of a century. This was surely an astonishing failure of the English legal system. What hope can there be for realistic, effective, prudent regulation of deeply complex fields such as artificial intelligence or quantum computing if the law pretends that computers are no more than simple machines? All I had to do was lift the lid and examine the mess underneath. All you have to do is click on the link and read my article, “The Law Commission and section 69 of the Police and Criminal Evidence Act 1984”.