29 mai 2018
The last decade has seen the emergence of significant technological developments in the legal sector and the introduction of artificially intelligent software in various guises. These are the first steps to automating the provision of certain legal services. Litigation is no different. The tools allowing the assisted investigation of large amounts of electronic information are already available and in use today. But what does the next decade hold?
Without the benefit of a time machine, we have trawled our imagination to pull together some ideas about what legal services in the disputes sphere might look like in ten years' time.
A standard step early on in any litigation is for a client to ask for a view on the merits of a claim or defence. That view is often crucial in the assessment of an early settlement, and the potential value of incurring further legal fees, etc.
The reality is that any assessment of the merits at an early stage is probably being undertaken before a full (or even just a thorough) review of the client's relevant documents has been completed. Existing Early Case Assessment is often conducted to a limited budget and so only on a limited number of documents.
What if you could ask your lawyer for a preliminary view of the merits, knowing that it would be prepared on the basis of an automated review of all the documents in your possession? Technology Assisted Review software already allows the automation of document review for disclosure after only a limited amount of training. We expect that in the near term that process will become even more automated and the software will be able to automatically highlight the documents which most support or undermine a claim or defence, allowing for a much more realistic view on the merits to be prepared right from the beginning.
During the course of any litigation, there is often the need to consider a large number of historic judgments and the extent to which the facts match or are similar to the present case.
The prospect of a legal robot to automate the process of legal research, including applying the relevant facts from the present case, to identify the key cases can't be too far from reality. An experienced practitioner in the relevant area of law would still be required to analyse the results and advise on the most appropriate course of action, but the process would be much quicker and less costly.
The ability to have an early and high-level assessment of the merits of a case against the entire bank of historic case law would be invaluable, both for clients and judges. Such a tool could also prove very useful during a hearing or trial when a judge may have specific queries as to whether there are any authorities to support a specific submission being made. The tool should make a search for such cases much quicker and more accurate. The scope of such a tool could also extend beyond the case law of England and Wales, such that it has a wider jurisdictional remit.
It is already possible for hearings such as Case Management Conferences to be held by telephone conference or to use video conferencing to allow witnesses to attend trial. Why not more?
With advances in global high speed internet and hologram technology video conferencing could soon be replaced by holograms, for added reality of the witness actually being "in the box".
In the longer term, why not dispose of the physical court room entirely for certain cases and instead hold hearings in the virtual world? With virtual reality advancing in leaps and bounds, we predict that we will have attended our first virtual hearing within the next 10 years. Such a practice would retain all the benefits of a hearing "in person" but afford substantial cost savings. International disputes where parties are spread across the world would be easier to schedule, less disruptive to clients and at lower costs. More flexible timetabling, as is currently more common in arbitrations, could also be introduced.
In complex commercial litigation, it is often necessary to call on the services of an expert with a very specific area of expertise. The expert is there to assist the court in understanding non-legal subject matters and to give their opinion.
Can it be far in the future that we no longer call upon a human expert for that assistance, but instead use sophisticated AI expert software, in whatever field an opinion is needed?
Knowing that the AI expert has considered and analysed all of the case information, together with all relevant industry texts and standards, might potentially make its opinion more credible than that one provided by a human expert. AI expert determination could become a new form of alternate dispute resolution in appropriate cases.
Such technology will not come without potential downsides. Cross-examining what is effectively a piece of computer software would not come without significant challenges. Questions would have to be precisely crafted, with an expert technician on hand to "translate" them. However, in certain types of cases, especially those where a joint expert could be appointed (e.g. valuers' negligence) there might be significant costs savings, and no risk of the expert being undermined during cross examination.
We already have electronic document bundles and automatic transcription of court proceedings. But might automatic fact checking of witness statements and expert reports become common? A natural and expected development in this area will be the automated real time matching of what is being said in court against the documents in the case.
Cases can be won or lost by a witness on the stand at trial. A judge must believe that a witness is being honest for what they are saying to be given any credibility. At present, the ability to challenge what is being said by a witness relies on the legal team instantly being able to recall facts and the content of specific documents. Those documents might only have been seen once or twice, often months in the past, when they held little significance to the case. All that can change in an instant at trial.
What if the live transcription was connected to software with access to the court bundles, and all the other documents in the case, with the ability to trawl the entire document universe for inconsistencies in an instant? Documents could then be brought to the attention of the barrister immediately to challenge the story being told. The cost of such a tool would be outweighed by the potential benefits at trial.
For low value disputes, it is often too expensive and time consuming to pursue much past an initial letter of demand. What if there was an option to issue a claim which would be dealt with entirely by AI and managed by the court?
The English civil courts already allow claims (via Money Claim Online) to be issued online and for the court to print and issue the relevant claim form for service on the defendant. A natural progression would be for court software to take a proactive approach and analyse the nature of any claim, ask for relevant information (be it narrative or documents), and then to reach a conclusion and issue a judgment. All that could happen without the claimant or defendant ever needing to step foot in a courtroom.
Not only would such an AI court process reduce the burden on the justice system, but also increase the ability of companies to make recoveries or engage in otherwise commercially uneconomic litigation.
par plusieurs auteurs