This post has been written by Associate Editor Rayan Bhattacharya.


The functioning and operation of the international arbitration industry, along with the legal industry in general, has constantly been impacted by global technological revolutions. In the 1990s, forecasts were made by the likes of Richard Susskind regarding the probable prevalence of e-mail, e-disclosures and e-filing technologies in the dispute resolution sector. A decade later, he was proven right. Today, there is certainly some merit in the calculated predictions that many industry veterans, Susskind included, have made about the impact of Artificial Intelligence (AI) and blockchain technology in dispute resolution. In fact, in 2017, he was brave enough to admit bluntly that the legal practice areas which refuse to embrace technological evolutions have a grim future ahead of them. At the same time, he also makes a convincing case for the opportunities awaiting sectors that do successfully embrace the technology.

The impact of these technologies on the legal industry as a whole has already been acknowledged at a global level. This can be evidenced in the fact that close to all global law firms, notably the UK Magic Circle and the US White-Shoe have strategically encouraged strategic incorporation of these technologies into their own practices. Even in emerging markets, rapid incorporations are being made in the legal industry. This is notably observed in China, where AI utility has not only become a rising norm in law firms but AI education a popular subject in law schools. In 2002, Constantine Partasides called arbitral secretaries as the “fourth arbitrator” outside the traditional three. This article will consider whether this classification can be taken a step further by including these technologies as the “fifth” candidate in an arbitration scenario and argue the conditions under which such an assertion will be possible. In order to demonstrate so, this article aims to analyse the opportunities and challenges posed by these technologies to the international arbitration sector in particular and seeks to identify the key functions of the sector that are likely to be impacted. The impact of each will be addressed separately in the following dedicated sections.

 Artificial Intelligence and International Arbitration

For the purposes of this analysis, AI in this legal context can be broadly defined as the technological process where large amounts of data are combined with intelligent self-learning algorithms. In essence, In 2018, the New York State Bar Association (NYSBA) published a special feature article on AI in International Arbitration, highlighting three potential benefits of the incorporation of the former in the latter. Firstly, AI can help enhance the appropriateness in the selection process of arbitrators themselves. Secondly, it can help reduce the time and cost of clerical legal processes. Thirdly, it is likely to reduce arbitral uncertainty. Each of these broad points is analysed in turn.

The first potential merit of appointing appropriate arbitrators stems from the longstanding issue of inherently biased, absent or incompetent arbitrators. The issue of bias alone has been repeatedly proven statistically as being a primary cause of incoherent arbitration decisions and judgments that often lead to no resolution of the dispute. Simply put, this issue is based on the premise that an arbitrator might have a preconceived notion on who should win a case before a fair evaluation of both parties’ cases. This results in a direct contradiction of Article 18 of the United Nations Commission on International Trade Law’s (UNCITRAL) Model Law on cross border arbitration. This rule calls for both parties in such a process to be treated in an equal manner, with Article 12 of the same Model Law allowing parties to challenge the bias of arbitrators. If estimated correctly, this merit can help implement the Screened Selection Process for Arbitrators advocated by the International Institute of Conflict Resolution & Prevention. This involves AI eliminating the unconscious bias of party-appointed arbitrators and also diversifying the pool of potential arbitrators. This is done with AI making sourcing and screening decisions based on the repeated tendencies of arbitrators while writing judgements, along with flagging and altering any such tendencies that may be deemed as biased. Similarly, while recruiting arbitrators, AI can be programmed to ignore any demographic data of candidates such as economic status or gender. Accordingly, any positive or negative associations made with any such demographic group will be neutralised. A move to ensure greater diversity will also be in line with the Equal Representation Pledge taken by the global Equal Representation in Arbitration Committee This pledge was the culminated effort of multiple law firms, arbitration academics and arbitration centres to increase the representation of women worldwide in the international arbitration sector. Hence, this solution will help reaffirm the classical notion of arbitration only being as good as the arbitrators themselves.

The second potential merit stems from the very common issue of lengthy legal procedures such as due diligence and contract review. The ability of AI platforms to hasten the duration of such procedures from the perspective of a lawyer has already been evidenced by the fact that most major global law firms have already adopted AI platforms for numerous internal tasks. A notable example of such a platform in recent years is Kira Systems’ contract review program. This system has been adopted by all five of the Magic Circle firms along with Latham & Watkins and Davis Polk being notable examples among the US White Shoe who have followed suit. However, with regard to international arbitration specifically, AI can also help recommend the ideal arbitration mechanism to pursue in case of a dispute in a contract. For instance, if parties want a unilateral arbitration agreement, an AI platform can automatically recommend the parties to choose London as the arbitral location instead of Paris, as the latter considers such forms of agreement as invalid.

The third potential merit stems from the possibility of AI making substantive legal contributions to the case of a party before the arbitration panel itself. The special feature of the NYSBA estimates that AI will soon be capable of predicting the chance of success of a case based on the nature of the arbitrators, likely costs to be incurred, historical success rate of arguments being used or even tendencies of a particular arbitration court. However, if objectively analysed, it can be reasonably argued that these estimations of the NYSBA fail to take into account the economics of availability of such AI platforms. The NYSBA’s argument is based largely upon the assumption that all parties in international arbitrations in the future will have equal access to such AI technology. However, in a scenario where only a few parties have the infrastructure to incorporate AI in their legal procedures, the rationale of AI making arbitration fairer and more productive for all parties will be defeated.

If universally implemented, the recommendations made by AI platforms on issues such as which arbitrators to choose, the chances of success or the history and the experience of the opposition and the arbitrators, have been tested to be highly accurate. The immediate implication of this is that the workload of junior lawyers who spend multiple hours on such logistical research is significantly reduced. In a way, it may also be argued that as AI platforms become more affordable, it will act as an equalising factor between parties who have the manpower to conduct large scale due diligence with junior lawyers and the ones who do not.

The basic principle behind his argument is the notion that robots do not work better in a big firm than a smaller firm, in the words of Peter Wallqvist, co-founder of a legal AI solutions provider called RAVN. This article’s argument is further supported by the confidence Wallqvist’s puts in smaller firms in becoming more open to AI solutions as prices go down and user-friendliness goes up.

Blockchain and International Arbitration

Historically, the call for the incorporation of blockchain technology in international arbitration popularly stemmed in response to traditional issues of cyber intrusions, confidentiality and efficiency in the arbitral process. While the term ‘blockchain’ itself may be defined in numerous ways depending on the context, for the purposes of this article, it may be defined as a distributed public ledger that contains the history and record of transactions and that can be accessed across a network.

The proposed marriage of blockchain with international arbitration has not been without scepticism. Doubts have been raised regarding the competence of public permission-less blockchain to ensure confidentiality of massive volumes of data. Hence, over the years, there has been popular consensus on the fact that the most suitable type of blockchain for international arbitration is to be a private and permissioned ledger. In simple terms, this means the blockchain has to ensure the confidentiality of the sensitive data involved in the proceedings, and at the same time ensure only pre-designated parties exercise control over the arbitration process and the blockchain (such as the arbitral tribunal, for instance). Using this category of blockchain as a working premise, the uses of blockchain in international arbitration are threefold as introduced at the beginning of this section.

Firstly, with regards to cybersecurity, the international arbitration sector has been no stranger to data breaches and intrusions. This was infamously evidenced in recent times in 2015 where the website of the Permanent Court of Arbitration in The Hague was hacked during a sensitive maritime arbitration between China and the Philippines. Similarly, Turkey, in the case of Libananco v Republic of Turkey,, had admitted of intercepting Libananco’s communications with its counsel under the justification of it being a part of a criminal investigation. More popularly, the arbitral tribunal in the case had to deal with newly available evidence through WikiLeaks. Law firms working along such arbitral proceedings have often faced breaches as well. A study of 200 such firms in early 2017 revealed that close to all firms had experienced breaches, with 40% of them not being aware of it when it occurred.

In this context, Deloitte has estimated that blockchains can help impede fraudulent activities and detect data tampering using its underlying characteristics of immutability, data encryption and operational resilience. Simply put, the fixed and electronically protected nature of a blockchain along with its ability to adapt to changing environments and conditions over a period of time help it fulfil the aforementioned functions. Furthermore, unlike other storage technologies like cloud computing, a blockchain does not have a single point of failure which decreases the chances of IP based distributed denial-of-service (DDoS) attacks. Additionally, the distributed nature of the ledger ensures that even if a significant portion of the blockchain network is under attack, it will continue to operate due to this very nature. Going beyond Deloitte’s analysis, it can also be reasonably argued that other common types of cyber-attacks such as drive-by attacks or man in the middle (MITM) attacks, among others, can be averted by AI. This is so as these attacks, in principle, aim to take control over legitimate sessions on a server or cause a shut down as well. This solution can potentially avert attacks at an earlier stage, by detecting and preventing the spread of the malware that facilitates the attack in the first place as per reports from Gartner Inc.

Secondly, the Queen Mary University Arbitration Survey of 2018 revealed that over 87% of parties to major international commercial arbitrations consider confidentiality as a necessity in such proceedings. Hence, in this context, it can once again be argued that private, permissioned blockchains ideally suit the needs of such parties. This will allow the blockchain to function like a private intranet page where only authorised individuals can access the protected data. This will be made possible using multiple virtual barriers such as passwords, voice and biometric recognition to ensure a truly multi-factor method of authentication. Hence, the risk posed to the confidentiality of proceedings, as evidenced in cases like Libananco, can be minimised.

Thirdly, the same 2018 Arbitration survey rated the cost and speed of arbitration as among the four worst qualities of arbitration, along with lack of sanctions during the arbitration process and lack of power with regard to third parties. This response itself is ironic considering the fact that arbitration as a procedure is touted to provide a more efficient alternative to traditional litigation. Once again, the question that arises is whether blockchain can help solve the issue of inefficiency in the industry. Yet again, there seems to be adequate evidence and consensus that it can. Blockchain technology has already begun to be seen as a tool deployed by select law firms to streamline back-office processes and in turn to increase speed and cut cost. A notable effort in this regard is Linklaters’ research report written along with the International Swaps and Derivatives Association on the potential applications of blockchain in the form of smart contracts in the legal sector. In a 2018 IBM report, it was gauged that the use of blockchain-based smart contracts can reduce the time consumed in dispute resolution by 75%. These logistical benefits made available by this technology at the micro-level in dispute resolution can in time be expanded to suit the needs of the international arbitration industry.


The popular upheaval regarding modernising the international arbitration industry is in many ways justified and necessary. Minor upgradations aside, the industry has fundamentally functioned the same way for over five decades. The potential benefits brought forward by both AI and blockchain is attractive but at the same time unclear and while the technologies themselves promise to act as cost-cutters and efficiency boosters, the larger economic practicality of procuring and incorporating such technology in the first place is yet to be gauged. If the premeditations regarding AI and blockchain do indeed prove to be true, and if the legal technology industry successfully manages to offer solutions that are cost-effective and accessible, then Constantine Partasides classification can be confidently taken a step further by calling AI and blockchain technology the “fifth arbitrators” in the modern arbitration industry.