In the last two decades, information and communication technologies (ICT) have progressed immensely, which has meant a significant increase in the level of innovation in products and services provided in many industries, including the legal sector. In this era of globalization and technological absorption, technical improvements have been introduced in the arbitration process, and the idea of arbitrators as replaceable by robots is looming, the most striking consequence of which is the emergence of many speculations.

Artificial Intelligence is emerging in  arbitration systems, but will the future of Artificial Intelligence become a realistic solution to disputes? The future is the answer.

Undoubtedly, some clear examples of the improvement in services provided by the arbitration community through the implementation of ICT are: videoconferencing, electronic disclosure, use of online platforms, cloud-based applications, and so on. However, to the extent of the consolidation of Artificial Intelligence, arbitration, as we know it today, will hardly exist in the future.

The major limitation of Artificial Intelligence remains its rebellious nature, which can make the use of the technology too wasteful and complicated. Moreover, although AI can automate low-level tasks, it is unlikely to supplant the manual arbitration of AI robots. In the end, all this adds to the degree of confidence needed to implement Artificial Intelligence in dispute resolution.


Artificial Intelligence in Modern Arbitration

Artificial Intelligence is a term coined for the general procedure of unifying large amounts of data with robust interactive processing systems and intelligent algorithms to drive automatic learning of software concerning patterns derived from the mass above information. However, the nomenclature “Artificial Intelligence” is often used imprecisely, as it encompasses various topics such as machine learning, cognitive computing, and natural language processing (Paisley et al., 2018).

The main difference between Artificial Intelligence and other automation and legal technology tools involves the ability to learn and develop along the way. Similarly, it is worth noting the existence of two main types of Artificial Intelligence mechanisms: rule-based learning and machine learning. Currently, most AI tools use machine learning, which is ideal for static and slowly varying scenarios.

At present, when there is a growing concern about the expenditure of resources and time in resolving litigation, AI has the potential to reduce the time and cost of resolving litigation and create incentives for early settlement. However, there are now concerns about the impact of this figure on decision-making and access to justice, depending on who has access to its benefits, the transparency, and control of arbitral data and algorithms, as well as the publication of awards and the potential risks to confidentiality and the protection of personal data, to name a few.

Even if it is inferred that an arbitration procedure may be carried out with the intervention of Artificial Intelligence, depending on the magnitude and qualification of such intercession, it is feasible to evaluate the precision and benefit of its application. With the fundamental purpose of making the arbitration process faster and more transparent, Artificial Intelligence can manifest itself in three different ways: as a support for the automation of procedural acts, as an interpretative tool for consultation, and as a judging entity in its own right.


Artificial Intelligence in the constitution of the Arbitral Tribunal

One of the first actions included in an arbitration process is the selection and appointment of arbitrators, who will hear and decide on the dispute. Thus, in particular disputes, the parties spend a considerable amount of time in a laborious process of choosing the arbitrators, which—with the help of Artificial Intelligence—can be mitigated to a great extent. At present, some platforms have been developed to help in this situation.

To illustrate this, specific mention is made of the Arbitrator Intelligence program, whose purpose is to act as a global information aggregator, collecting qualitative and quantitative data on legal professionals and users in terms of critical elements for the choice of arbitrators. Through this platform, the selection of arbitrators is optimized by analyzing a vast amount of information comprised of the history of potential candidates and their linkage with the nature of each case in question (CIAR Global, 2020).

At the same time, this type of dynamic not only implies a clear saving of time in the arbitration process but also provides transparency to a crucially important arbitration stage, such as the selection of the tribunal members. Above all, to counteract the implications of the phenomenon of “double-hatting,” which implies the risk of conflict of interest arising from the simultaneous concurrence of certain legal professionals as representatives of parties in certain arbitration proceedings and as arbitrators in others (Fierro Valle, 2014).


Artificial Intelligence as an interpretative consulting tool

Among the various forms of application of Artificial Intelligence in arbitration is the processing of data and precedents for the proposal to the judges of possible decision designs. However, this interpretative function of consultation certainly does not involve the direct resolution by Artificial Intelligence but the formulation of patterns, whose function is to serve as a suggestion to the arbitrators to elaborate their opinion.

From now on, many software and applications can be pointed out to fulfill the abovementioned purpose. An example is the “Arbilex” initiative, which uses Artificial Intelligence to issue a coherent predictive analysis regarding possible suitable arbitration results. Likewise, one can denote “Premonition” as one of the vastest litigation databases in the world; “Context” as a means capable of evaluating millions of pieces of arbitral jurisprudence for the structuring of possible awards; and “Kira” as automated learning and Artificial Intelligence software aimed at the identification and interpretation of contracts and documents.

In short, today, the alternative of Artificial Intelligence has been installed at the service of the arbitration process as a mechanism that can be used by the parties involved in it. However, participation in this particular means of dispute resolution has been limited, at least, to acting as a fourth party and not as a kind of immediate judge of disputes. Otherwise, this paradigm could mutate over time.



Even though the technology is constantly growing, formulating such a radical alteration as the attribution of the decision-making function to an Artificial Intelligence system in arbitration entails a series of unknowns and hypothetical but imminent disadvantages that are difficult to solve.

In short, Artificial Intelligence will play a crucial role in arbitration as a means of conflict resolution shortly. The benefits provided by this technological tool are too great to be missed. However, this reality can only be magnified by materializing guarantees on the accuracy and scope of machines to deal with all those unpredictable but common circumstances that abound in arbitration proceedings.


Michelle Bernier is an attorney specializing in international law and commercial law. She is currently studying Master of Laws and International Business with a double degree from the Universidad Internacional Iberoamericana in Mexico and the Universidad Europea del Atlántico. She is also a part of Students for Liberty’s inaugural cohort of Fellowship for Freedom in India.