Diales Compendium Issue 3 - Flipbook - Page 13
ISSUE 3
COMPENDIUM
ARTIFICIAL INTELLIGENCE IN DISPUTES
It is widely agreed that the courts in general are
overburdened, and most disputes take a long time to proceed
and resolve. Some papers published concerning machinedeveloped advanced AI propose replacing humans in the
legal system, similar to Online Dispute Resolution (ODR)
where the complaints can be automatically logged online,
and question whether in the future the AI algorithm will
examine the evidence and resolve the issue without any
human interaction.
An AI system is likely to be more emotionally objective and
give what is considered to be a strictly ‘logical’ outcome,
but may miss the subtle nuances that become apparent in a
court room and may not be able to accurately replicate the
court’s decision-making methodology and reasoning behind
the damage awards. It is also difficult to envisage how the
cut and thrust of the cross-examination process can be
mimicked by the AI process.
In the AI decision-making world, the human interaction,
emotion and body language are missing. Some might prefer
to avoid the hostile environment of an open court by having
an AI resolution, but many will favour the human interaction
and emotion it gives.
Although AI may have its limitations when it comes to the
judgment process, it can be very helpful in managing large
amounts of case data that requires review. Recent advances
in AI technology mean that lawyers and practitioners have
more sophisticated AI tools at their disposal to rapidly process
data and identify relevant information, such as through
a sophisticated dictionary that can perform comparative
language checks (synonyms, antonyms, etc.,). In the future,
it may be possible to narrow the issues by AI processes
identifying the points of agreement and disagreement from
the data set. However, the outcome of the AI will only be
as good as the data set that it is given, and it may not be
able to identify missing information that could be critical
to the dispute. The security of the system must also be a
primary concern to prevent hacking with a desire to skew the
outcome.
It is possible that if the necessary technology can be
developed, AI could become useful for low value disputes or
disputes requiring a rapid turnaround.
When a subject matter of the dispute is highly technical, AI
may or may not be able to partially fulfil the role of arbitrators
with an appropriate degree of expertise, but the outcome,
decision and awards remain the most critical part of the
process that currently cannot be relied upon.
VIRTUAL AND AUGMENTED REALITY
There may also be a place in the court or arbitration system
for an immersive, 3D environment, such as VR (Virtual Reality)
or AR (Augmented Reality), which can be a more engaging
way to communicate complex information. Presently, 2D
visualisations are generally used. An immersive virtual 3D
reality may be more useful in visualising complex issues and
explaining them to the court or the tribunal, such as a 3D
model of a bridge collapse or other events that have multiple
causes. Platforms like Metaverse or Second Life, etc., provide
realistic environments for users where they can immerse
themselves using VR headsets.
The 3D immersive experience may be the next practical
stage in adoption of computer-based technology if this has
an advantage over the 2D visualisation process. Although
Diales Technical has not yet ventured into using VR such as
Metaverse, it is proud to be one of the firms using advanced
technology to gather facts and evidence using in-house
developed technology to speed up the process. Diales
employs computer-based technology in many forms and
systems to assist our experts in dispute resolution services.
In summary, assistive technology can provide critical tools
to gather and process information for dispute resolution
purposes. Assisted technology is currently a useful solution
for discovery, helping to reduce costs relating to document
searches related to discovery. In time, AI technology may
develop and be used for judicial decision-making as legal
teams and clients gain confidence in the quality of the
decisions through a process of trial and error and appeals that
determine to what extent the AI judgment is determinative
and binding.
However, AI is currently not mature enough to engage with
real-life problems and resolve disputes without any human
input. The most difficult issue to resolve using AI is the
outcome of the dispute, for which currently no technology
seems to exist and is likely to require some sort of selflearning software similar to DeepMind.
13