In 2020, because of the pandemic, British schoolchildren were unable to sit public examinations, and so were given grades based on their past work. For lazy but bright students who had planned to idle all year before cramming furiously in the last two months, this policy will have been as welcome as a hog roast at a Bar Mitzvah. They were to be judged on their past performance, not their future excellence. Then, to make matters worse, the Department for Education adjusted students’ grades using a simple algorithm that took into account the performance of previous students at the school from which each pupil came. This had the (apparently unintended) effect of downgrading 39 per cent of results, with poorer pupils seeing the greatest downward adjustment because schools in less prosperous areas had previously underperformed. Three days later, young people took to the streets of London with placards reading ‘Fuck the Algorithm.’
-Susskind, Jamie. The Digital Republic (p. 83). Bloomsbury Publishing
At the moment the paper based County Court system seemingly stands on the point of collapse: telephone enquiries are routinely sourced to centralised call centres with 90 minute waiting times, correspondence to courts is not answered for weeks, it is impossible to get a face to face appointment with a member of court staff and documents lodged at court routinely end up in a Black Hole, lost by some strange force of gravity and doomed never to reach a judge. Even a small claim takes on average 51 weeks to get to a hearing.
It is therefore with some scepticism that one contemplates the introduction of a new integrated digital justice system, which is meant to abolish paper and use modern technology to bring together civil family and tribunal cases in a digital process which will run from the cradle to the grave. Will this prove to be the latest public sector IT White Elephant, or will it really prove to be as ground-breaking as its proponents would have us believe?
On 28th April 2022, the Judicial Review and Courts Act 2022 received the Royal Assent and became law. Chapter 2 of that Act provides the raw sinews for the digitisation of civil and family justice in England and Wales. It provides the broad framework by which the resolution of disputes is to be moved online and resolved digitally, using newly created Online Procedure Rules and provides a statutory basis for an Online Procedure Rule Committee to come into existence. This body will have extremely broad powers indeed to set rules for the online resolution of disputes. What is envisaged has been described as a three layered “funnel”: the first layer will be a website and an app to which any would be claimant can go to find out how to progress a claim of any kind. Claimants will then be signposted to a series of pre-action portals and ombuds processed to identify and seek to resolve their claim. Any claim that is not resolved within the appropriate pre-action space will have a data set that will then be transmitted to the third layer of the funnel which is a court based online justice process, such as Online Civil Money Claims and Damages Claims online. That may lead to a court hearing, it may lead to a remote court hearing by video, or it may lead to a “digital paper” based determination.
So far so good, but what is envisaged is far more than a digital filing system. Instead the system is meant to be “intelligent” so that it can identify the issues between the parties, it can determine how the result should be resolved, and it can make interventions including suggesting online or telephone or in person mediations, or early neutral evaluations or the use of so called artificial intelligence to suggest outcomes, and potentially at some point, to determine outcomes, subject to review or appeal by a human judge.
The potential gains from increase in the speed with which justice is administered, the potential for saving costs and the potential for efficiency in getting to the heart of a dispute to resolve it, are breath-taking in scope. The dangers are equally significant depending on how much more is built into this system whether at its inception or later in its evolution.
Let me explain what I mean. If the procedures include algorithms, then those algorithms may have unfavourable, if not unforeseeable consequences. Algorithms tend to suffer from the “neutrality” fallacy: that they produce a just result built on analysis of data patterns. The data does not lie. But the prejudices and biases of society arising from language, sex, and ethnicity will simply follow through into the data. Correlation is different from causation. There is no such thing as artificial intelligence, only machine learning by systems which work by processing huge quantities of data and detecting patterns. Suppose such a system hypothetically, takes into account that more fraudulent road traffic insurance claims emanate from certain post codes than others? Would it be legitimate for the system to take that into account when determining the result or suggesting the disposal of a specific road traffic accident claim? What if the residents of that postcode are predominantly of a particular ethnic origin?
Moreover, whatever rules the Online Procedure Committee creates will then have to be implemented by means of computer code, which will underpin apps and online portals by which the new system will be operated. The danger of such an approach is clear and manifest. If the Online Procedure Rules provide for something to be done within 28 days, but the coding relevant to that step inadvertently provides for something to be done within 21 days, then 21 days will be the requirement the parties practically face online. In effect, the coders will determine the legal rules, notwithstanding what the Online Procedure Rules Committee determines they should be. I am not suggesting that this will be done deliberately, but it is often hard enough for qualified lawyers to determine what procedural rules mean and how they should be applied.
The questions which have not even been asked yet, still yet answers provided, is what role for lawyers still exists, when you can formulate a claim by simply opening an app on your smartphone, which will then formulate it for you, identify the issues, and attempt to get you to settle it, with either gentle or stern prodding with offers of mediations and early neutral evaluations. The question that also falls to be considered is that even if lawyers are to be retained, how are they to be remunerated and what role do costs have in a digital system? If the system itself, has identified the issues and gathered the data in support of the claim and in defence of the claim, then how is the work done by the lawyers to be evaluated and paid?
Looking into the longer term future for the civil justice system, when considering those disputes which do make their way to litigation, some commentators, have suggested that the rise of the block chain, the internet of things, the relentless accrual of data running from smart contracts, to CCTV footage, to smartphones that track your location to the metre and the minute, and the digital footprint left by emails and voice calls will mean that disputes over fact, including in the personal injury space and the clinical negligence world will become redundant, as everything will be “known”.
Such futurism may yet prove misplaced. In any number of science fiction novels and films from years gone, by the 1990s there were meant to be flying cars and easy travel to Mars. What will probably retard such developments, is the slow and partial pace of technological change for most of the population: look on the roads around you and see how many cars are 10 or 15 years old, with engineering that may seem old hat, but is adequate for its owners purposes. Also note the millions of people in the United Kingdom who do not have access to the internet or even a bank account. Also note that despite extremely widespread CCTV coverage in the UK in 2022, how infrequently there is useable footage which appears in any personal injury dispute. Perhaps also foreseeably on the horizon are the problems of proof that will be generated by ostensibly reliable digital records: a credit card payment may not prove that someone was in London on a given date and time, but merely that their credit card may have been cloned. At the current time “deepfake” video technology seems to be used simply to create videos for TikTok, but could be used to manufacture evidence, to “prove” that someone was present (or not) at the scene of an accident.
Something must be done. All too often reform in the past, has taken the form of the shuffling of the deckchairs on the deck of the Titanic. The possibilities of a truly digital online system have been missed, either because the technology was not there, the funding was not there, or the imagination was not there. But if I had to cast the auguries for the digital future, I would suggest it is not going to be a digital superhighway free from bumps and potholes.
This article was written for that excellent publication, PI Focus, and its fresh and new format launched this month: please see the entire issue, including this article, here: