Various technological developments and their intersection with arbitration have recently been subject to vigorous debates. Particularly with regard to the COVID-19 pandemic, there has been a proliferation of smart technologies, which are geared towards increasing efficiency and quality of arbitrations. One such technological development sparking discussion has been Artificial Intelligence (AI). Several AI-based legal tech tools in the market have proved useful in terms of saving time and cost of arbitrations.

The essence of commercial arbitration is rooted in the efficient adjudication of legal disputes. With machine learning processes of the AI, these tasks can be readily automated. This begs the question whether AI could serve a bigger purpose in assuming important roles and tasks undertaken by arbitrators.

Merits of the AI-arbitrators  

International arbitration tends to be expensive and lengthy. It appears that the AI-based arbitrations could be able to change that. With an ability to process vast amounts of data within seconds at a fraction of the costs otherwise incurred, AI is highly attractive to parties concerned about matters of efficiency and spending.

Secondly, the paramount duty of an arbitrator is to be independent and impartial. As such, if AI-arbitrations were to become common practice, it could counteract biases and minimize concerns of undue influence. Similarly, there would be less chance of a conflict of interests as AI, being a tool for automation simulated through a computer program, is capable to adjudicate upon the dispute objectively.

Not only are conscious and unconscious biases, otherwise inherent to human nature, absent in AI software, human arbitrators are also more prone to offering scholarly opinions on an issue that may be central to a case and thus likely to stir up conflict.

Drawbacks of the AI-arbitrators

Adjudication is a long-established, adversarial progress, whose many benefits include finality as well as the safeguard to ensure parties due process under the law. Public adjudication also offers the advantage of human interaction and the ability to evaluate expressions of witnesses more sensitively, which a software might interpret in a number of different ways.

Moreoever, the software is only as good as the data that it is fed. If used in a dysfunctional and biased manner, algorithms could lead to discriminatory results. Similarly, limited entries will render selective information outcomes. Therefore, while able to offer a ‘binary response based on probabilistic inference [AI] may obscure many controversies under the guise of objective analysis.1

Furthermore, the practice of arbitrators draws on a combination of factors, inter alia experience, emotion and empathy. Deciding a case in arbitration, is not merely based on inductive, but deductive reasoning, i.e. giving consideration to specialized domain knowledge, expertise and practical understanding.2 Further, beyond lacking human discretion that is vital to dispute resolution or the cognitive abilities central to the legal decision-making process, AI would also deprive court users of their right to be provided with reasons for the outcome of their case.3

It is for all of the aforesaid factors that a technological advancement such as AI, neither substitute nor render instituionalised court-based adjudication redundant.


While it seems improbable that AI will replace human arbitrators in the near future, the latter could undoubtedly use its software to their benefit in addressing time and costs concerns. AI-based platforms such as Opus 2, Luminance, Litigate AI, Ross Intelligence, etc. have already made consireable strides in transforming the practice of arbitration – offering solutions that are datadriven, more expedient and may narrow the possibility for error.

Nevertheless, the notion of arbitration being conducted by human arbitrators is here to stay. While the democratization of substantive arbitration data through AI technology certainly offers considerable advantages, it is difficult to fathom how human arbitrators, at present, could be replaced particularly when considering matters such as the sanctity of party-nomination procedures. Neverthleless, if proving to serve commercial business interests and current technologies continue to be improved and built upon, the prospect of AI-based arbitrations can no longer be excluded and may in fact be inevitable.


1 Chauhan, A. S.  (2020) Future of AI in Arbitration: The Fine Line Between Fiction and Reality. Kluwer Arbitration Blog. Available via: [accessed 10.12.2020].

2 Ibid.

3 Ibid.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.