Richard Clark writes on “Will AI Change Dispute Resolution?”

Art Resolve’s Hetty Gleave writes on AI and art: how recent court cases are stretching copyright principles in the Art Newspaper
14 April 2023
Art Resolve’s Hetty Gleave contributes her expertise on how to protect artworks from theft
17 November 2023

Originally published on

The short answer is yes.    

Just as AI is already playing a growing role in the business of law firms and the practising of law, it will inevitably become a part of the dispute resolution arena. The more difficult question is how and over what period of time might we see really significant changes? 

AI has already been used to assist parties in settling their disputes. However, in my own area of large commercial disputes, its role is currently limited and not widespread. So where might we see further inroads of AI and what are its limitations as an effective means of ADR? 

Data-driven outcomes

Everyone knows that data is intrinsic to AI. It is astonishingly good at processing and analysing huge amounts of data. This makes it potentially very good at doing what lawyers spend a good deal of their time doing – applying the law to the facts of their client’s case and predicting legal outcomes. 

So, in the field of formal dispute resolution, AI is highly likely to play a significantly increasing role in predicting outcomes – how courts and tribunals are likely to resolve disputes. Theoretically, this may lead to more and earlier settlements before matters reach the court or tribunal because each side is more likely to reach the same conclusion on the merits of the case.  

However, the flip side of AI’s formidable big data analysis capabilities is its voracious appetite for data. It only works well when there is a complete (or at least very substantial) data set available to analyse. The size and availability of reported and decided cases in most developed jurisdictions is exactly what AI needs to deliver reliable and accurate outputs and predictions. AI is likely to get better and more impressive at doing this as it learns and accesses more and more available data.  

Many lawyers would correctly argue that there is far more to advising clients than analysing big data and predicting outcomes. Nevertheless, it is hard to imagine a future in which AI does not play an increasing and significant role in the area of dispute resolution just as it will in so many other areas of legal practice.    

AI and mediation

The ability of AI to produce valuable analytics and predictions in the field of litigation and disputes will become more relevant to the mediation process, just as it will to advising clients and participating in non-mediated negotiations. However, will AI ever replace the role of a human mediator completely? 

Most commercial lawyers and mediators would answer ‘no’ to this question and for very credible reasons. 

First, unlike reported judicial decisions, there is no available big data cataloguing the outcomes of commercial mediations. So, the ability of AI to learn from the experience of actual mediation outcomes is very limited. Of course, data sets exist but most remain private and known only to the mediator and the parties who are generally under continuing obligations to keep the outcome confidential. It follows that unless and until mediation outcomes are so widely available that AI has sufficient data to predict the outcome of a mediation, it is unlikely to render the process of human mediation obsolete or redundant.    

One of the big advantages of mediation over ending up in court is privacy and confidentiality. This does not sit well with AI and there would need to be all kinds of protections built into any future AI-led mediation processes which would in turn tend to reduce the effectiveness of the AI.   

Secondly, mediation is very different from a formal dispute process such as a trial or arbitration. Of course, the legal merits are an important constituent of any mediation but they are far from the only consideration in what is always a complex and multifactorial process.    

Emotions, feelings, body language, nuance of language and human presence are all crucially important in most commercial mediations and a skilled mediator is constantly listening, accessing and leveraging these intangible factors in helping the parties navigate pathways to resolution and settlement. Of course, this is not to say online mediation is ineffective – it can be highly effective. However, it is interesting to note in passing that, following the necessity of online mediation during the pandemic, many large commercial disputes have returned to being mediated in person.    

These considerations probably mean that (at least in the short to medium term), the vast majority of clients, lawyers and mediators are unlikely to embrace AI-led mediation, especially in large commercial disputes. That is not to say it will not increasingly be a very useful tool in analysing disclosure documents and data relevant to the dispute. However, in the longer term, things may be very different.   

The future 

History suggests that one would be unwise to predict where we might be if we take a look further into the future (and that might mean less than a decade given the rapidity of progress and AI learning). Nevertheless, it seems unlikely the pace of development will reduce or that genies will be going back into bottles.   

It is likely that more sophisticated AI will become highly effective at reading human emotions, body language and nuances as well as, and probably better, than humans. Very few areas which we currently regard as uniquely human are likely to be out of reach to future generations of AI and this includes mediation.  

Perhaps the pace of change and the key limiting factors will not be the technology itself but our willingness as humans to embrace and adopt it. Major scientific and technological advancements always encounter a period of initial scepticism and resistance but this inevitably gives way to acceptance and adoption. This process is helped by generational socialisation as the young who have known a world only with AI enter the workforce and further drive adoption and acceptance. The history of email and social media are obvious examples of this process.   

Finally, I asked AI for its opinion. In answer to the question: will AI replace human mediators? Chat GPT came to pretty much the same views as I have expressed in this short discussion. It concluded by saying:    

“While AI technologies can enhance the efficiency and effectiveness of mediation processes, human mediators are likely to remain essential for the foreseeable future. The combination of human expertise and AI tools has the potential to greatly enhance the practice of mediation, providing better outcomes for parties involved in disputes.” 

So perhaps my days of writing articles are numbered but I can assume I shall still be mediating disputes for a few years yet.  

Richard Clark 

June 2023