Artificial Intelligence in International

Arbitration: A Step Too Far?

Annabelle O. ONYEFULU*

The emergence of the global pandemic has propelled the increased usage of artificial intelligence
(Hereinafter referred to as ‘AI’) in the workplace. Technological advancement in the legal
environment, evident in such functionalities as digitalization, blockchain, and machine learning,
amongst others, have significantly affected international arbitration. The application of these
technologies may promote efficiency, cost, and speed on the one hand, whilst perhaps, violating
the current regulatory frameworks in international arbitration on the other. Accordingly, this
research article undertakes a brief analysis of the future of international arbitration vis-à-vis the
introduction of new technologies, including an evaluation of the possibility of replacing human
arbitrators with AI arbitrators. Attention is paid to the public policy requirements and ethical
issues that may arise therefrom.
Keywords: Artificial Intelligence (AI), International Arbitration, AI Arbitrators, Technology, Public
Policy, Ethical Issues

1 INTRODUCTION
There is no universally accepted definition of the term ‘artificial intelligence (AI)’;
it has been variously defined by several authors, scholars, scientists, and mathema-
ticians, amongst others. John McCarthy defined the term as ‘the science of
creating, developing and use of intelligent machines, software, data processing
machines amongst others to analyse macro data using algorithms’.1 Ryan Calo
defined it as ‘a process which creates a machine that consists of a well-designed and
analytic approximation of human and animal cognition’.2
The use of AI-based technologies in the twenty-first century, especially in our
everyday lives, cannot be overemphasized. Following the emergence of the global
pandemic in 2019, technological advancement gained speed as it became evident

*
LLB, BL, LLM, ACIARB UK, Doctoral Researcher, Oxford Brookes University.
Email: [email protected].
1
John McCarthy, What Is Artificial Intelligence? (2007), http://jmc.stanford.edu/articles/whatisai/whati
sai.pdf (accessed 14 Feb. 2022).
2
Ryan Calo, Artificial Intelligence Policy: A Primer and Roadmap (2017), https://ssrn.com/abstract=
3015350 (accessed 14 Feb. 2022).

Onyefulu, Annabelle O. ‘Artificial Intelligence in International Arbitration: A Step Too Far?’. Arbitration:
The Int’l J. of Arb., Med. & Dispute Mgmt 89, no. 1 (2023): 56–77.

© 2023 Chartered Institute of Arbitrators

 A STEP TOO FAR? 57

that business transactions, legal cases, arbitral proceedings, academic classes, and
even yoga lessons can all be done virtually. The development of science and
technology applications during the pandemic has significantly affected international
dispute resolution, thereby resulting in some foreseeable transformations in inter-
national arbitration. The possibility of including technology in international arbi-
tration has engaged scholars for quite some time. Whereas some scholars have
embraced the fact that parties, counsel and arbitrators are in dire need of such
inclusion, others are vehemently opposed to this.3 The fact remains, however, that
humans are constantly looking for ways to make their lives somewhat easier, hence
the constant need to develop and advance in technology.
Foreseeable transformations proposed by AI include (but are not limited to)
the application of algorithms in the arbitral process and the replacement of human
arbitrators with AI arbitrators.
The replacement of human adjudicators with AI adjudicators already exists in
some countries.4 Some jurists have contended that this is indeed a welcome
innovation, as AI Arbitrators can produce expert decisions in human-readable
form which will be transparent and devoid of bias.5 Some other jurists are against
this as they contend that AI judges will lack the most important element in their
decision which is judicial persuadability.6 This debate will be extensively discussed
in this research article.

2 PRACTICABILITY OR OTHERWISE OF REPLACING HUMAN

ARBITRATORS WITH AI ARBITRATORS
AI software that is adjudicatory in nature is widely proposed, and to some extent
already in use. Thus the jobs of arbitrators are threatened.7 The possibility of AI
arbitrators replacing human arbitrators has been a topic of contention by several
scholars during the past few years.8 Eugene Vodoh, a renowned professor and

3
H. Eidenmüller & F. Varesis, What Is an Arbitration? Artificial Intelligence and the Vanishing Human
Arbitrator (2020), https://ssrn.com/abstract=3629145 (accessed 14 Feb. 2022).
4
For Example, Estonia has developed an AI judge which is capable of adjudicating small claims cases
with a monetary jurisdiction of about 7,000. Also, China has digitalized courts and AI judges as the
head adjudicators. See Eric Niiler, Can AI Be a Fair Judge in Court? Estonia Thinks So, WIRED MAG.
(25 Mar. 2019). See https://www.wired.com/story/can-ai-be-fair-judge-court-estonia-thinks-so/
(accessed 26 Apr. 2022). See also Beijing Internet Court Launches AI Judge, China Daily (2019), http://
www.chinadaily.com.cn/a/201906/28/WS5d156cada3103dbf1432ac74.html (accessed 26 Apr. 2022).
5
Richard E. Susskind, Expert Systems in Law: A Jurisprudential Approach to Artificial Intelligence and Legal
Reasoning, 49 Mod. L. Rev. 168, 173 (1986).
6
Andrew C. Michaels, Artificial Intelligence, Legal Change, and Separation of Powers, 88 U. Cin. L. Rev.
1083 (2020).
7
Christine Sim, Will Arbitration Take over Arbitration?, 13(1) Asian J. Int’l Arb. 1–14 (2018).
8
Young-Yik Rhim & Kyung Bae Park, The Applicability of Artificial Intelligence in International Law, 12(1)
J. East Asia & Int’l L. (2019), http://dx.doi.org/10.14330/jeail.2019.12.1.01 (accessed 14 Feb. 2021).
58 ARBITRATION: THE INT’L J. OF ARB., MED. & DISPUTE MGMT

scholar of law, has posed a question: ‘If AI advances to the point where it can
adequately mimic judicial opinion writing, should we accept an AI judge?’9 This
raises a bone of contention in the entire legal community, including practitioners
of forms of dispute resolution such as arbitration. Some jurists welcome the
prospect of AI judges with open arms, while others are vehemently opposed to
such an innovation.
Professor Vodoh himself has expressed his interest in welcoming this change
in the legal community. He expressly stated that; ‘If a system proposes change and
it is seen by members of that system as a form of development to that system, it
should be ACCEPTED rather than resisted’.10 Some scholars have criticized this
opinion, contending that a judgment is said to be reasoned if there is an element of
judicial persuadability which is derived from factual legal arguments presented by
the parties.11 They point out that an AI judge or arbitrator lacks human feelings
and cannot be persuaded, and that no matter how the software or programme is
designed there would necessarily be a huge problem in deciphering the factors that
would persuade an AI judge or arbitrator.12 Put another way, this means that one
of the elements in determining who can be an arbitrator is the ability of the person
(or robot) to be judicially persuaded. Thus, this research article proceeds by asking
the million-dollar question; who can be an arbitrator?

3 WHO CAN BE AN ARBITRATOR?

Upon the commencement of a dispute in international arbitration, parties or their
appointing authority are at liberty to choose an arbitrator or arbitrators subject to
the dictates of their arbitral contract.13 The court also has the inherent power to
make such appointment in certain circumstances – for example, where a party to a
dispute is a state entity14 or if parties or their appointing authority fails to appoint
an arbitrator.15 Parties or the appointing authority can choose to appoint a natural
person or an arbitral institution to preside over the dispute.16 Several procedural
laws and conventions (and some scholars) have expressly or by implication pro-
vided that arbitrators must be natural persons. Article 1450 of the Code of Civil
Procedure of France, for example, expressly defines an arbitrator as a natural person
having the full legal capacity to render adjudicatory powers on an arbitral dispute

9
Eugene Volokh,Chief Justice Robots, 68 Duke L.J. 1135, 1138 (2019).
10
Ibid.
11
Michaels, supra n. 6.
12
Emily Berman, Individualized Suspicion in the Age of Big Data, 105 Iowa L. Rev. 463, 502 (2020).
13
Article 11(2) UNCITRAL Model Law.
14
Article 13(4) ICC Rules.
15
Section 18 Arbitration Act 1996.
16
Article 11 UNCITRAL Model Law.
 A STEP TOO FAR? 59

and who is duly appointed by parties, their appointing authority or the court.17
Article 3 of the Scottish Rule of Arbitration defines an arbitrator as ‘a sole
arbitrator who must be an individual18 and such person must pass the eligibility
test which includes: necessary qualification, impartiality and independence’.19
Article 11 of the United Nations Commission on International Trade Law
(UNCITRAL Model law) impliedly provides for arbitrators to be natural persons.
It provides that a person shall not be enstopped from representing parties as an
arbitrator by virtue of his/her nationality.20 It is in doubt if an intelligent machine
can have a nationality. According to the law of most states, there are only two
types of legal entities with nationalities: natural persons and juridical persons.21
Nonetheless, parties or their appointing authority may appoint arbitrators subject
to the contract. The parties can state the procedure for appointing arbitrators and
the number of arbitrators to be appointed. It is important to note that if this
procedure is not adhered to, a party may request that the court steps in.22 It is
worthy of note that the requirement for appointment of arbitrators under the
UNCITRAL model law remains: qualifications, impartiality and independence.23
Furthermore, Article 13 of the International Chamber of Commerce (ICC)
rules also impliedly provides for natural persons as proposed arbitrators, because the
nationality and residence of a proposed arbitrator act as criteria for the appointment
of an arbitrator.24 The court, under this rule, in appointing a sole arbitrator must
appoint an arbitrator with a different nationality and/or residence from the parties
unless the parties expressly agree otherwise.25 The ICC rules also state the basic
requirements to be qualifications, impartiality and independence.26 Article 6 of the
London Court of International Arbitration (LCIA) rules also refers to the nation-
ality of arbitrators as a criterion for appointment as the presiding arbitrator27 or
emergency arbitrator.28 This inherently means that an arbitrator is a person in law
capable of having a nationality and is eligible (that is, qualified according to the

17
Article 1450 Code of Civil Procedure of France.
18
Article 3.1 Scottish Short Form Arbitration Rules 2012, s. 8 Arbitration (Scotland) Act 2010.
19
IBA Guidelines on Conflicts of Interest in International Arbitration, Part I (2014) n71, Art. 3.8
Scottish Short Form Arbitration Rules 2012.
20
Article 11 UNCITRAL Model Law.
21
Elvia Arcelia Quintana Adriano, The Natural Person, Legal Entity or Juridical Person and Juridical
Personality, 4 PSJL 5 (2015). See https://elibrary.law.psu.edu/cgi/viewcontent.cgi?article=1117&con
text=jlia (accessed 21 Mar. 2022).
22
Article 11(4) UNCITRAL Model Law.
23
Article 12 UNCITRAL model Law.
24
Article 13 ICC Rules.
25
Article 13(5) &(6) ICC Rules.
26
Article 14 ICC Rules.
27
Article 6 LCIA Rules.
28
Article 9B LCIA Rules.
60 ARBITRATION: THE INT’L J. OF ARB., MED. & DISPUTE MGMT

dictates of the parties agreement), impartial, and independent. An intelligent

machine does not appear to possess all of these qualities.
Nevertheless, it is worthy of mention that in all the arbitration regulations
mentioned above, not even one expressly disqualifies an AI-based tool from acting
as an arbitrator.29 There is also no provision for the unenforceability of an award
rendered by an AI Arbitrator under the New York Convention.30 The point that
has, however, been argued by several jurists is that this convention was ratified in
1958 and at that time the draftsmen could not possibly have envisaged the use of
AI Arbitrators.31 This is of course true, but nonetheless the position remains that
by virtue of Article II (2) of the New York Convention, a court can enforce an
award rendered by a corporate or legal entity.32 If AI tools are at some future date
ratified as legal entities capable of rendering an award, then their awards will fall
under this subsection for enforcement.33

4 PRACTICABLE GROUNDS FOR CHALLENGING THE

ENFORCEABILITY OF AN ARBITRAL AWARD RENDERED
BY AN AI ARBITRATOR
4.1 ATTESTATION
Many jurists have contended that a possible ground for challenging the enforce-
ment of an arbitral award rendered by an AI arbitrator is attestation of an award.34
They contend that most national and international laws provide that one of the
elements that validates an arbitral award, or court judgment is the attestation, or
signature of the arbitrator(s).35 Article IV(1)(a) of the New York Convention
expressly states that an arbitral award shall be recognized and enforced by the
court if a party seeking its enforcement submits a duly authenticated award to the
court, or a certified copy thereof.36 Article 31 of the UNCITRAL Model Law
provides that an arbitral award shall be in writing and signed by the arbitrators.37

29
Sim, supra n. 7, at 3.
30
Ibid.
31
Ibid. Gülüm Bayraktaroğlu-Özçelik & Ş. BarışÖzçelik, Use of AI-Based Technologies in International
Commercial Arbitration 9 (2021), https://ejlt.org/index.php/ejlt/article/view/814/1033 (accessed 14
Feb. 2021).
32
Article II(2) New York Convention provides that ‘The term “arbitral awards” shall include not only
awards made by arbitrators appointed for each case but also those made by permanent arbitral bodies to
which the parties have submitted’.
33
Sim, supra n. 7, at 3.
34
Bayraktaroğlu-Özçelik & BarışÖzçelik, supra n. 31, at 9.
35
Ibid.
36
Article IV(1)(a) New York Convention 1958.
37
Article 31 UNCITRAL Model Law.
 A STEP TOO FAR? 61

Thus, since an AI tool cannot produce a written signature, it can be inferred that
an AI arbitrator cannot render a final and binding arbitral award.38
Conversely, some other jurists are of the opinion that having human arbitra-
tors alongside an AI arbitrator may solve the problem of admissibility of the arbitral
award, and issues resulting therefrom such as authentication and signature.39
Adopting this system will, however, pose unanswered questions such as; who
becomes the umpire if there are two human arbitrators and an AI arbitrator.40
Where there are conflicts in decisions or opinions of arbitrators, how will the
opinion of the AI arbitrator be explained?41 If two AI arbitrators are appointed and
one human arbitrator, will the authentication of only one arbitrator suffice?42
Many other such questions will arise. However, the creation of a distinctive
mark or stamp for AI arbitrators may solve the issue of signatures.43
Consequently, in the event that a human arbitrator is paired up with an AI
arbitrator, signature might not necessarily be an issue.

4.2 SELECTION OF ARBITRATORS IN ACCORDANCE WITH THE ARBITRAL AGREEMENT

Awards rendered by an AI arbitrator may be challenged on grounds stated in

Article 34(2)(a)(IV) of the UNCITRAL Model Law and Article V(1)(d) of the
New York Convention, which provides that an award can be challenged if the
arbitral tribunal is not composed of arbitrators agreed to by parties, or is in
significant contravention of the law of the seat of arbitration. Parties involved
in an arbitral proceeding have the autonomy of selecting their arbitrators44 as
well as determining how the proceedings will be conducted.45 This autonomy
is, however, subject to their arbitral agreement. When the arbitral agreement
states that natural persons should be appointed as arbitrators and there is a
subsequent appointment of an AI arbitrator, an unsatisfied party can raise that
as a ground for challenging the award under Article 34 of the UNCITRAL
Model Law and Article V of the New York Convention. Moreover, even if
parties to an arbitral proceeding expressly agree to the appointment of AI
arbitrators and the law of the seat of arbitration (lex arbitri) only permits the

38
Bayraktaroğlu-Özçelik & BarışÖzçelik, supra n. 31, at 9.
39
Epaminontas E. Triantafilou, Non-lawyer Arbitrators and the Deliberative Balance of International Arbitral
Tribunals, TDM 7 (2016).
40
Sim, supra n. 7, at 4.
41
Ibid.
42
Ibid.
43
Ng (Huang Ying) & Benedetti del Rio, When the Tribunal Is an Algorithm: Complexities of Enforcing
Orders Determined by a Software Under the New York Convention, 121 Kluwer 34 (2019).
44
Article 11 UNCITRAL Model Law.
45
Article 19 UNCITRAL Model Law.
62 ARBITRATION: THE INT’L J. OF ARB., MED. & DISPUTE MGMT

appointment of a natural person, the law of the seat shall prevail by virtue of
Article 34(2)(a)(IV) UNCITRAL Model Law.46 That is, an award rendered by
an AI arbitrator duly appointed by parties can be challenged if such appoint-
ment is contrary to the lex arbitri. However, it is important to note that, for a
party to successfully raise this issue as a ground for challenging the arbitral
award, it must be successfully raised during the early stages of the
proceedings.47

4.3 CONTRAVENTION OF THE LAW OF THE SEAT OF ARBITRATION

An award rendered by an AI arbitrator can be challenged by virtue of Article 34 of the

UNCITRAL Model Law and Article V of the New York Convention on the ground
that the composition of the arbitral tribunal is in significant contravention of the law of
the seat of arbitration, as mentioned above. However, most relevant national and
international legislation does not explicitly either bar or permit the use of AI
arbitrators,48 and some jurists contend that since this lacuna exists it may be resolved
in favour of AI arbitrators.49 There is nothing in the UNCITRAL Model Law barring
the use of AI arbitrators nor is there anything in the New York Convention that bars
the enforcement and recognition of awards rendered by AI arbitrators.
Nevertheless, it is important to note that some legislation are now being enacted,
which expressly bars the use of AI arbitrators. For example, by virtue of Article 1450
of the Civil Code of Conduct of France, only natural persons can be appointed as
arbitrators. Article 7 of the Turkish International Arbitration Law contains the same
stipulation. Some other laws have stated that ‘individuals or persons’ who have rights
and capacity should be appointed as arbitrators.50 However, interpretation of the issue
discussed in this subsection in the future may become a herculean task, given that the
European Parliament has passed a resolution which addresses robots as electronic
people,51 thus suggesting that AI arbitrators may fall into the category of ‘person’.

46
Article 34(2)(a)(IV) UNCITRAL Model Law.
47
Bayraktaroğlu-Özçelik & BarışÖzçelik, supra n. 31, at 11.
48
Nairobi Centre for International Arbitration, Artificial Intelligence (AI) in International Arbitration:
Machine Learning (2021). See https://ncia.or.ke/wp-content/uploads/2021/08/ARTIFICIAL-
INTELLIGENCE-AI-IN-INTERNATIONAL-ARBITRATION.pdf (accessed 22 Mar. 2022).
49
Ibid.
50
Article 3.1 Scottish Short Form Arbitration Rules 2012, s. 8 Arbitration (Scotland) Act 2010.
51
European Parliament Resolution of 16 Feb. 2017 with recommendation to the commission on civil
law rules on robotics (2015/2103(INL)).
 A STEP TOO FAR? 63

5 PUBLIC POLICY REQUIREMENT

Public policy has no specific definition. It has been described by a scholar as
principles of morality and justice which citizens of a nation consider to be of national
interest, and it varies from state to state.52 Public policy is a concept which may
change from time to time in a society.53 Article 34 (2)(b)(ii) of the UNCITRAL
Model Law54 and Article V (2)(b) of the New York Convention55 provide that a
court may set aside an award if it is contrary to public policy. Whilst what can violate
public policy is not expressly spelled out in neither the UNCITRAL Model Law,
nor the New York Convention unlike its specification in other public international
regulations,56 the construction or interpretation of a manifest contravention of public
policy can be inferred through a case-by-case analysis.57 It has been contended that
an arbitral award rendered by an AI arbitrator may contravene several public policy
grounds.58 For example, where the national legislation of the seat of arbitration
expressly provides that an arbitrator must be a natural person, any award rendered by
an AI arbitrator is in contravention of the public policy of the national legislation.59
In addition, the recognition and enforcement of an award rendered by an AI
arbitrator which is contrary to the lex arbitri will also amount to a contravention
of public policy under Article V(2)(d) New York Convention.60
Furthermore, an award can be set aside, and recognition and enforcement
denied if a party satisfies the court that the arbitrators were not independent or
impartial in their dealings.61 This can serve as a public policy ground.62 It has been
contended by several jurists that the use of AI arbitrators in rendering decisions in
arbitral proceedings will be flawed with data bias.63 AI arbitrators can only function

52
Andrew I. Okekeifere, Public Policy and Arbitrability Under the UNCITRAL Model Law, 2 Int’l A.L.R.
70–77 (1999).
53
Ibid.
54
Article 34(2)(b)(ii) UNCITRAL Model Law.
55
Article V (2)(b) New York Convention.
56
Regulation (EC) No 593/2008 of the European Parliament and of the Council of 17 Jun. 2008
(‘Rome I’), Art. 21; Regulation (EU) No 1215/2012 of the European Parliament and of the Council
of 12 Dec. 2012 on jurisdiction and the recognition and enforcement of judgments in civil and
commercial matters (recast), Art. 45(1)(a), Art. 46; Swiss Federal Act on Private International Law, Art.
27(1); Turkish Private International Law Act amongst others.
57
Bayraktaroğlu-Özçelik & BarışÖzçelik, supra n. 31, at 12.
58
Ibid.
59
Ibid.
60
Eidenmüller & Varesis, supra n. 3.
61
IBA Guidelines on Conflicts of Interest in International Arbitration, supra n. 19.
62
M. V. Benedetelli, Human Rights as a Litigation Tool in International Arbitration: Reflecting on the ECHR
Experience, 4 Arb. Int’l 59 (2015).
63
J. J. Angwin et al., Machine Bias (2016), https://www.propublica.org/article/machinebias-Risk-assess
ments-in-criminal-sentencing (accessed 22 Mar. 2022). See also A. Caliskan et al., Semantics Derived
Automatically from Language Corpora Contain Human-Like Biases, 356 Sci. 183–186 (2017), see also
Bayraktaroğlu-Özçelik & BarışÖzçelik, supra n. 31, at 13.
64 ARBITRATION: THE INT’L J. OF ARB., MED. & DISPUTE MGMT

when the algorithms have been fed enough data by humans.64 This data can be
discriminatory in nature as it may contain elements likely to affect race and
gender.65
Another potential ground for challenge of an award or denial of recognition and
enforcement of an award on the ground of public policy, may be linked to the
absence of a reasoned decision in an arbitral award.66 By virtue of Article 32(2) of the
UNCITRAL Model Law, an award shall be valid if it contains a reasoned decision
of an arbitral tribunal unless the parties expressly agree that the reasoning should not
be stated, or the award is based on an agreement entered into by the parties.67 This
means that, in jurisdictions where the national legislation provides for the reasoning
of an award as a strict requirement for validity, the absence of such will amount to a
violation of public policy.68 Algorithms lack feelings and empathy, and certainly
cannot make use of discretion,69 all of which are components of a reasoned award.
Justice is much more than a processed macro data and a set of precedent decisions; it
is distinctly not a well-constructed algorithm.70 It is a human virtue which is
characterized by a range of factors such as evaluating and analysing key facts and
legal arguments, application of the law of the state, and the use of discretion to arrive
at what a tribunal feels right and just for all parties involved.71
Nevertheless, awards rendered by AI arbitrators might be welcomed by parties
and not challenged on grounds of public policy if it is contended that there are
certain advantages that might be associated with an AI arbitrator acting as an
adjudicatory authority. Such advantages may include the following: duration of
proceedings may cease to become worrisome to parties as proceedings may be
determined in a matter of months as opposed to years in the human arbitrator
regime.72 Some scholars have widely embraced this concept and stated that the:
dissatisfaction and grievances occasioned by resort in arbitration as a medium of dispute
settlement by parties which is associated with time, cost and the nonchalant behaviour of
members of the arbitral community will be greatly reduced by the introduction of
technology to this system and making such technology affordable to parties.73

64
Ibid.
65
Ibid.
66
M. Scherer, International Arbitration – How Artificial Intelligence Will Change Dispute Resolution, Austrian
Y.B. Int’l Arb. 514 (2019).
67
Article 32(2) UNCITRAL model law.
68
Scherer, supra n. 66.
69
Sim, supra n. 7.
70
Nairobi Centre for International Arbitration, supra n. 48.
71
Ibid.
72
Ibid.
73
Paul Cohen & Sophie Nappert, The March of the Robots (2015), http://globalarbitrationreview.com/
article/1080951/the-march-of-the-robots (accessed 22 Mar. 2022). See also Jack Wright Nelson,
Machine Arbitration and Machine Arbitrators (2016), http://www.youngicca-blog.com/machinearbitra
tion-and-machine-arbitrators/ (accessed 22 Mar. 2022).
 A STEP TOO FAR? 65

Algorithms may also be developed which are capable of learning and applying
precedent decisions in a matter of minutes, thereby under some circumstances
rendering a better decision than would human arbitrators.74
It is, however, worthy of note that several jurists have contended that algo-
rithms are generally susceptible to inaccuracy, and that because the underlying
software is kept secret there is no means of reviewing decisions.75 This will deter
the strict adherence to decisions rendered by AI arbitrators.76

6 IMPLICATION OF REPLACEMENT OF HUMAN ARBITRATORS

WITH AI ARBITRATORS IN THE FUTURE OF ARBITRATION
The implication of AI-based technologies for the future of international arbitra-
tion cannot be gainsaid, as it promises to provide solutions to inherent issues such
as duration of proceedings, cost, lack of legal consistency, and speed, amongst
others, thereby favouring parties and their counsel as well as arbitrators.77 AI-
based tools are also being designed to render predictive decisions so parties
embarking on an arbitral proceeding are likely to know the outcome. This
may play a major role in deciding whether parties decide to negotiate or not.
The replacement of human arbitrators with AI arbitrators, however, is a game
changer in the arbitral community78 as intelligent machines may take over the
adjudicatory process. Although no such algorithm has yet been developed, this
innovation will certainly change the system of dispute resolution in the coming
years.79
Nevertheless, albeit a welcome innovation, the resultant effects of this
replacement of the arbitral system will include unemployment, as AI is targeted
to take over the jobs of arbitrators.80 AI arbitrators will be designed in such a
way that their work requires little or no human intervention thereby requiring
fewer humans and causing unemployment.81 Moreover, there is also the pos-
sibility of rendering an erroneous award tainted by gender or racial bias. It has
already been established that AI arbitrators can only function when fed with

74
Nairobi Centre for International Arbitration, supra n. 48.
75
Sim, supra n. 7.
76
Ibid.
77
Ibid.
78
Bayraktaroğlu-Özçelik & BarışÖzçelik, supra n. 31.
79
Ibid.
80
Arnav Joshi, Replacing Arbitrators With Artificial Intelligence in International Arbitration (2020). See https://
www.thearbitrationworkshop.com/post/replacing-arbitrators-with-artificial-intelligence-in-interna
tional-arbitration (accessed 22 Mar. 2022).
81
Anandaday Misshra, Artificial Intelligence(AI) and Its Effects on Arbitration (2020). See https://taxguru.in/
corporate-law/artificial-intelligence-ai-effects-arbitration.html (accessed 23 Mar. 2022).
66 ARBITRATION: THE INT’L J. OF ARB., MED. & DISPUTE MGMT

data. There have been incidents where existing algorithms have been reported
to make judgments in a discriminatory manner due to the information that was
fed to them. A notable algorithm of this type was the Amazon recruitment
database, which was proven to be gender-biased thereby employing more men
than women.82
Furthermore, in cases where an established pattern exists that involves a
particular set of people dominating a branch of arbitration, an AI arbitrator is
likely to rule in favour of that set in any future dispute without looking into the
merits of the case due to the data that was fed to it.83 For example, in international
investment arbitration where investors have in most cases won cases over host
states, an AI arbitrator may in any future dispute award the case to the investor
without looking at the merits of the case.84 This is because an algorithm’s output
largely depends on its input; this is derived from the concept ‘garbage in, garbage
out’.85
Another major legal impact an AI arbitrator could have on arbitration is the
issue of absolute immunity granted to the arbitrator by statute.86 By virtue of
Article 41 of the ICC rules, an arbitrator is granted with immunity to render
awards, only limited only by decisions made in bad faith.87 This means that, if it
is established that a human arbitrator has rendered one or several awards in bad
faith, he or she will be liable and may face disciplinary action which may result
in dismissal from office. However, where the arbitrator is a machine arbitrator
or AI arbitrator, the ultimate question will be: Who is responsible? The
developer of the algorithm? The person who fed data to the algorithm? The
machine itself?88
Nonetheless, there seem to be positive impacts of an AI arbitrator replacing a
human arbitrator. Scholars have opined that the unpredictability associated with
arbitral awards will decrease as AI tools are developed to predict cases. Moreover,
AI arbitrators will greatly help in solving problems of prolonged duration of
proceedings, cost, and legal inconsistencies associated with many if not most
awards.

82
Jeffrey Dastin, Amazon Scraps Secret AI Recruiting Tool that Showed Bias Against Women (2018). See
https://www.reuters.com/article/us-amazon-com-jobs-automation-insight/amazon-scraps-secret-ai-
recruiting-tool-that-showed-bias-against-women-idUSKCN1MK08G (accessed 23 Mar. 2022).
83
Jean R. Sternlight, Creeping Mandatory Arbitration: Is It Just?, 57 Stan. L. Rev. 1631, 1650 (2005).
84
Rhim & Bae Park, supra n. 8, at 10.
85
Oscar Tang, Think Arbi: Will Artificial Intelligence Help or Harm Arbitration? (2021). See http://arbitra
tionblog.kluwerarbitration.com/2021/12/22/think-arbi-will-artificial-intelligence-help-or-harm-arbi
tration/ (accessed 23 Mar. 2022).
86
Mahnoor Waqar, The Use of AI in International Arbitration (2021), https://papers.ssrn.com/sol3/papers.
cfm?abstractid=3931233 (accessed 23 Mar. 2022).
87
Article 41 ICC Rules.
88
Waqar, supra n. 86.
 A STEP TOO FAR? 67

7 IMPLICATIONS OF AI ARBITRATORS IN AFRICA

Arbitration in Africa is not new as over thirty-three out of fifty-four countries in Africa
are signatories to the New York Convention.89 There has been wide adoption and
ratification of the UNCITRAL Model Law for the countries that are signatories to the
New York Convention and actively practice arbitration.90 Most national laws that
provide for arbitration in Africa neither expressly prohibit nor expressly permit the
appointment of AI arbitrators.91 For example, the CAJAC (China–Africa Joint
Arbitration Centre), an arbitration centre in Johannesburg, South Africa established
by China and African stakeholders, in its rules adopts the definition of ‘arbitrator’
according to the UNCITRAL Model Law.92 It is however, pertinent to note that the
Uniform Arbitration Act enacted in 2017 by the Organization for Harmonization of
Business Law in Africa Organisation pour l’Harmonisation en Afrique du Droit des
Affaires (OHADA)93 – formed by seventeen countries in order to enhance the
investment climate in Africa – expressly provides that an arbitrator must be a natural
person.94 Therefore, if parties in a proceedings appoint AI arbitrators, such appoint-
ment will be in contravention of the lex arbitri and will not stand.95
Nonetheless, parties from African countries that are not signatories to UAA
but have adopted the UNCITRAL Model Law may appoint AI arbitrators as there
is no prohibition of such appointment. The appointment of AI arbitrators in
Africa, however, may encounter such challenges as the following:

(1) Lack of Information: It has been established that the success of an

algorithm depends on the information ‘fed’ to it.96 Although it is
enshrined in the law of most African countries, arbitration as a
mode of dispute resolution is not widely embraced.97 Parties prefer
to refer disputes to traditional courts or litigation. The resultant
effect is that few awards have been rendered by arbitral tribunals in
Africa.98 This common response to arbitration is associated with the

89
Sadaff Habib, The Use of Artificial Intelligence in Arbitration in Africa – Inevitable or Unachievable? (2020),
IBA. See https://www.ibanet.org/article/E62B06F6-7772-458A-A6E7-1474DB7136B5 (accessed 23
Mar. 2022).
90
Ibid.
91
Lagos State Arbitration Act 2004.
92
See China–Africa Joint Arbitration Centre at http://www.shiac.org/CAJAC/index_E.aspx (accessed
23 Mar. 2022).
93
Ibid.
94
Article 5 UAA 2017.
95
Article 11 UNCITRAL Model Law.
96
Waqar, supra n. 86.
97
Ibid.
98
Ibid.
68 ARBITRATION: THE INT’L J. OF ARB., MED. & DISPUTE MGMT

fact that arbitral proceedings are usually prolonged and expensive,

often beyond parties’ means. It is also pertinent that the publication of
an award rendered by an African arbitral tribunal is usually resisted,
following the universal doctrine of confidentiality.99 Efforts are,
however, being made in Africa as to the publication of certain aspects
of the proceedings. For example, by virtue of Article 21(6) of the
Arbitration Foundation of South Africa rules, the publication of
protocols in remote arbitration hearings is permissible.100
(2) Lack of trained AI personnel: AI expertise in Africa is quite
thin as there are very few AI professionals or knowledgeable
teachers. AI research projects are poorly funded (or non-existent
in many countries) and eligible institutions are poorly equipped.101
Governments have shown little or no interest in funding AI
projects as they see no revenue value.102
(3) Maintenance of AI arbitrators: Even though the introduction of
this type of technology in Africa may be of great significance as it will
enhance speed and lead to shorter duration of arbitral proceedings, cost,
award prediction and much more,103 it may equally lead to a number of
foreseeable difficulties, in particular the lack of maintenance of the
robots as there is a massive problem with power supply in many
African countries and a lack of experts in the field to update the software
when required.104 And the problem of unemployment is also present,
as it was reported in 2018 by scholars at Oxford University that, in
Africa as elsewhere, AI is likely to take over jobs in future as developers
devise algorithms with the skills of an arbitrator.105
(4) Accessibility of AI arbitrators: The issue of accessibility to this
innovation may be a herculean task in Africa. Some scholars have
contended that the cost of acquiring a high-tech AI arbitrator
(robot) will be very expensive and unaffordable to Africa’s impo-
verished countries.106

99
Dolling-Baker v. Merrett (1990) 2 ALL ER 890.
100
Article 21(6) Arbitration Foundation of South Africa.
101
N. Butcher, M. Wilson-Strydom & Baijnath, Artificial Intelligence Capacity in Sub-Saharan Africa:
Compendium Report (2021), AI4D. See https://africa.ai4d.ai/wp-content/uploads/2021/03/AI4D-
Report%E2%80%94AI-in-SSA.pdf (accessed 23 Mar. 2022).
102
Ibid.
103
Waqar, supra n. 86.
104
Ibid.
105
Ibid.
106
Kathleen Paisley & Edna Sussman, Artificial Intelligence Challenges and Opportunities for International
Arbitration, 11(1) N.Y. Disp. Res. Law. 35–40 (2018).
 A STEP TOO FAR? 69

8 ETHICAL CONSIDERATIONS
There are a number of ethical concerns associated with the use of AI in interna-
tional arbitration that go beyond the replacement of human arbitrators with AI.
These concerns include but are not limited to the following:

8.1 FAIRNESS OF PROCEEDINGS

There have been major concerns as to the fairness of proceedings that will be
conducted by an AI arbitrator.107 This is mainly associated with the quality of
output of an AI tool which has been established to be totally dependent on the
input. That is, an AI tool can only render an award based on the data that has been
fed to it. Some scholars have contended that there is a possibility of data bias, and
this may likely affect gender and race discrimination.108 This suggests that the
fundamental requirements for appointment of arbitrators (i.e., impartiality and
independence)109 may be breached. This is complicated in that parties in interna-
tional arbitration cannot generally appeal the decision of an arbitrator as the
decision of an arbitrator is final and binding except in limited circumstances and
only on questions of law and not fact.110 This is a major ethical issue because, in
the event that an award is unfair to a party based on racial or gender bias grounds,
the party cannot appeal the decision except in these limited circumstances. Some
legislation provides that parties can only challenge the recognition of the award.111
This is still not a satisfactory situation, as a number of unanswered questions arise,
such as: Who can be held responsible, the developer of the algorithm? The person
who inputted the data? The AI arbitrator itself?112

8.2 VALIDITY OF THE AWARD

Another ethical concern to be inferred from the replacement of human arbitrators

with AI arbitrators involves the validity of an award. A primary element of an
award is the decision of the arbitral tribunal, which depends on reasoning.113 An
award rendered by an AI arbitrator has been argued to lack reasoning for its
decision.114 This is associated with the fact that an AI arbitrator lacks emotions

107
Ibid.
108
Angwin et al., supra n. 63.
109
IBA Guidelines on Conflicts of Interest in International Arbitration, supra n. 19.
110
Article 35 UNCITRAL model law, s. 69 English Arbitration Act 1996.
111
Article 35 UNCITRAL Model Law.
112
Waqar, supra n. 86.
113
Article 32(2) UNCITRAL Model Law.
114
Sim, supra n. 7.
70 ARBITRATION: THE INT’L J. OF ARB., MED. & DISPUTE MGMT

and therefore cannot provide the reason for its decision. In the case that a party is
not satisfied with the decision of the AI arbitrator and wants to challenge the
award, this will lead to a major issue as the court will not be able to determine the
reasoning behind any decision made by the AI tool.

8.3 CONFIDENTIALITY
One of the major attractions of arbitration as a mode of alternative dispute
settlement is the privacy and confidentiality associated with the process.115 In the
event of any development and subsequent use of an AI arbitrator, there might be
demands for the publication of arbitral awards thereby defeating parties’ interest.

8.4 SECURITY

With the high rise of cybercrime currently taking place in the world, an ethical
concern may arise as to security measures put in place to protect AI-based tools from
hacking, spam, malicious viruses and other such criminal attacks.116 Inadequate
security measures can manipulate AI Arbitrators into making biased decisions.
It is worthy of note that the European Commission in 2018 set up a group, the
European Commission’s High Level Expert Group on Artificial Intelligence,117
whose primary responsibility is to evaluate prevalent ethical issues stemming from
the use of AI and set regulatory guidelines. These guidelines stipulate policies to
ensure that AI tools are human-centric and in compliance with the Charter of
Fundamental Rights of the European Union.118 They focus on the principle of
justice, requiring judgments to be free from bias or any form of stigmatization.119

9 CONCLUSION
This discourse has demonstrated and critically analysed the consequence of the use
of AI-based technologies in international arbitration and the possibility of replacing
human arbitrators with AI Arbitrators.

115
Dolling-Baker v. Merrett, supra n. 99.
116
James Kwan, James Ng & Brigitte Kiu, The Use of Artificial Intelligence in International Arbitration: Where
Are We Right Now?, 22 Int’l A.L.R. 19–26 (2019).
117
The European Commission’s High Level Expert Group on Artificial Intelligence, Draft Ethics
Guidelines for Trustworthy AI: Working Document for Stakeholders’ Consultation (2018). See https://ec.
europa.eu/digital-single-market/en/news/have-your-say-european-expert-group-seeks-feedback-
draft-ethics-guidelines-trustworthy (accessed 25 Mar. 2022).
118
Paragraph 4 of the Draft Ethics Guidelines for Trustworthy AI: Working Document for Stakeholders’
Consultation.
119
Ibid.
 A STEP TOO FAR? 71

This research article analysed the practicability or otherwise of the replace-

ment of human arbitrators with AI arbitrators. Although the algorithm for this has
not yet been fully developed, the article analysed some national legislation on
arbitration as well as the UNCITRAL Model Law as to the possibility of this novel
innovation to the arbitral community.
From this analysis, it can be determined that no national legislation expressly
bars the appointment of AI arbitrators. National laws and the UNCITRAL Model
Law expressly or by implication state that natural persons only can be appointed as
arbitrators. Although this can be contested on a general principle of contract law
which states that silence does not constitute acceptance (by the authority of McGlove
v. Lacey120), nonetheless it is a known principle of law that what is not expressly
admitted is deemed not to be denied and facts not denied are deemed to have been
tacitly admitted. This has been the bone of contention in this debate, especially given
that Articles II and V of the New York Convention do not expressly deny the
recognition and enforcement of awards rendered by AI arbitrators. Some scholars
have argued for the irrelevance of this contention, indicating that the appointment of
AI arbitrators as at 1958 when the convention was ratified was never in contempla-
tion as the draftsmen could not possibly have foreseen such a possibility.
With the emergence of the global pandemic in 2019, AI applications gained
popularity, and the use of AI technologies in arbitration now seems more plausible.
It can be regarded a welcome innovation as it has benefits such as; enhancing
speed, shortening the duration of proceedings, reducing cost, and in general
promoting the efficiency of the system. Major concerns with the appointment of
AI arbitrators, as discussed throughout this research article, include lack of infor-
mation, data bias, no reasoned decision, and weakening of due process.
Public policy concerns have also been raised in this research article as, by virtue of
Article V(2)(b) of the New York Convention, this objection could serve as a ground
for challenging the recognition and enforcement of an arbitral award rendered by an
AI arbitrator. Public policy grounds can stem from a number of reasons, including
composition of an arbitral tribunal. If the relevant national legislation expressly pro-
vides for the appointment of only natural persons as arbitrators and parties nevertheless
appoint AI Arbitrator, this is a contravention of the national legislation and therefore a
valid public policy ground. Another potential public policy ground could be the
indeterminacy of the reason for the decision. A strict element of a valid award is the
inclusion of the reason for arriving at such a decision and award. This is a human
element, as a human arbitrator can be persuaded to render an award after a careful
examination of evidence, legal arguments, and discretion. One scholar has referred to
judges as flesh and blood persons as opposed to a robots that merely recites the data

120
McGlove v. Lacey (1968) 288 F. Supp.
72 ARBITRATION: THE INT’L J. OF ARB., MED. & DISPUTE MGMT

which has been fed to ‘it’.121 It is, however, uncertain if an algorithm can be
developed to have emotions and function with discretion.
In any event, it can be agreed at present that public policy serves as a major
roadblock to the smooth entering into force of the machine arbitrator’s regime.
Recognition and enforcement of awards rendered under such a regime may still
encounter major difficulties under the New York Convention.
This study concluded by examining the ethical considerations occasioned by
the adoption of an AI regime. A party could contend that the use of AI arbitrators
could lead to severe unfairness and partiality in the case. This may be as a result of
data bias, and subsequently not being able to appeal will deprive that party of the
fundamental right to a fair hearing.

10 RECOMMENDATIONS
10.1 AMENDMENTS OF LEGAL INSTRUMENTS

The amendment of the New York Convention and re-enactment of national and
international instruments expressly enabling the use of AI-based technologies, appoint-
ment of AI arbitrators, and recognizing awards rendered therein will be one of the first
steps in accepting this new system. It should be noted that discussions are currently
ongoing as to the amendment of the New York Convention.122 This amendment will
ensure that most public policy grounds for challenging an award or enforcing an award
will be silenced so the appointment of an AI arbitrator will not contravene the dictates of
any legal instrument. Moreover, given that the major attractions of arbitration are the
autonomy and flexibility of parties to be able to control the system, mandatory aspects of
legislation which may hinder this flexibility should be abolished.

10.2 PROVISION OF SUFFICIENT DATA

Arbitral institutions should start publishing awards or making such publication

accessible when demanded in an anonymized form.123 This will assist in curbing
erroneous decisions when the new AI system is in place. This is necessary because
the AI arbitrator must be furnished with all relevant precedent decisions so it can
apply them to cases.

121
Ryan Calo, Robots as Legal Metaphors, 30 Harv. J.L. & Tech. 209, 218–219 (2016).
122
Tirado, Acevedo & Cosio, Draft Convention on the International Enforcement of Arbitration Agreements and
Awards 310 (2019).
123
ICC, Note to Parties and Arbitral Tribunals on the Conduct of the Arbitration Under the ICC Rules of
Arbitration (2019), https://iccwbo.org/content/uploads/sites/3/2017/03/icc-note-to-parties-and-arbi
tral-tribunals-on-the-conduct-of-arbitration.pdf (accessed 25 Mar. 2022).
 A STEP TOO FAR? 73

10.3 FUNDING OF AI RESEARCH PROJECTS IN AFRICA

Adequate funding and deployment of trained personnel to assist in the implemen-

tation of this new system in developing countries especially in Africa will yield
positive results as people will be trained on how to make use of this system.

10.4 EASY ACCESSIBILITY TO THESE TECHNOLOGICAL TOOLS BY LESS DEVELOPED

COUNTRIES

The less developed countries or countries that cannot easily afford to make use of
these technological tools should be assisted in one way or another. There should be
uniform access to these tools by every country that has adopted and is practicing
arbitration, as failure to do so will lead to ground for challenging the transparency
and nondiscrimination of this process.

10.5 FEATURE SELECTION

There should be some kind of manual or regulation guiding the kind of data that
will be ‘fed’ to AI arbitration software ensuring that it is devoid of discrimination
or bias, with features that ensure it can always be updated with the latest precedents
and laws.

11 BIBLIOGRAPHY
11.1 ARTICLES

J. Angwin, et.al, Machine Bias, Pro Publica (2016).

E. A. Adriano, The Natural Person, Legal Entity or Juridical Person and Juridical
Personality, PSJL 5 (2015).
N. A.-S. Ahmed Mohammad Al-Hawamdeh, The Effects of Arbitrator’s Lack of
Impartiality and Independence on the Arbitration Proceedings and the Task of Arbitrators
Under the UNCITRAL Model Law 64–66 (2018).
M. Bendedetell, Human Rights as a Litigation Tool in International Arbitration:
Reflecting on the ECHR Experience, Arbitrational Int’l 59 (2015).
E. Berman, Individualized Suspicion in the Age of Big Data, Iowa L. Rev. 503
(2020).
R. Calo, Artificial Intelligence Policy: A Primer and Roadmap 3 UBLR 2(2017).
J. Dastin, Amazon Scraps Secret AI Recruiting Tool that Showed Bias Against
Women (2018).
74 ARBITRATION: THE INT’L J. OF ARB., MED. & DISPUTE MGMT

H. Eidenmüller & F. Varesis, What Is an Arbitration? Artificial Intelligence and the

Vanishing Human Arbitrator (2020).
S. Habib, The Use of Artificial Intelligence in Arbitration in Africa – Inevitable or
Unachievable? (2020)
J. N. James Kwan, The Use of Artificial Intelligence in International Arbitration:
Where Are We Right Now?, Int’l A.L.R. 19–26 (2019).
A. Joshi, Replacing Arbitrators With Artificial Intelligence in International Arbitration
(2020).
J. McCarthy, What Is Artificial Intelligence (2007).
A. C. Michaels, Artificial Intelligence, Legal Change, and Separation of Powers, U.
Cin. L. Rev. 1083 (2020).
A. Misshra, Artificial Intelligence (AI) and Its Effects on Arbitration (2020).
P. C. Nappert, The March of the Robots (2015).
Ng (Huang Ying) & Benedetti del Rio, When the Tribunal Is an Algorithm:
Complexities of Enforcing Orders Determined by a Software Under the New York
Convention 34 (Kluwer 2019).
A. I. Okekeifere, Public Policy and Arbitrability Under the UNCITRAL Model
Law, Int’l A.L.R. 70–77 (1999).
Gülüm Bayraktaroğlu-Özçelik & Ş. BarışÖzçelik, Use of AI-Based Technologies
in International Commercial Arbitration (2021).
Young-Yik Rhim & Kyung Bae Park, The Applicability of Artificial Intelligence
in International Law, J. East Asia & Int’l L. (2019).
M. Scherer, International Arbitration – How Artificial Intelligence Will Change
Dispute Resolution, Austrian Y.B. Int’l Arb. 514 (2019).
C. Sim, Will Artificial Intelligence Take Over Arbitration?, Asian J. Int’l Arb. 1–14
(2018).
J. R. Sternlight, Creeping Mandatory Arbitration: Is It Just?, Stan. L. Rev. 1650
(2005).
S. P. Strezhnev, Affiliation Bias in Arbitration: An Experimental Approach, SSRN
Elec. J. (2016).
R. Susskind, Expert Systems in Law: A jurisprudential Approach to Artificial
Intelligence and Legal Reasoning, 47 Mod. L. Rev. 168, 173 (1986).
O. Tang, Think Arbi: Will Artificial Intelligence Help or Harm Arbitration? (2021).
A. Tirado, Draft Convention on the International Enforcement of Arbitration
Agreements and Awards 310 (2019).
E. E. Triantafilou, Non-lawyer Arbitrators and the Deliberative Balance of
International Arbitral Tribunals, TDM 7 (2016).
E. Vodoh, Chief Justice Robots, Dukel. J. 1135 (2019).
M. Waqar, The Use of AI in International Arbitration (2021).
11.2 BO
 A STEP TOO FAR? 75

11.2 CASES

Dolling-Baker v. Merrett, 890 (ALL ER 1990).

McGlove v. Lacey, 288 (F. Supp 1968).

11.3 LAWS

Article II New York Convention 1958.

Article IV(1)(a) New York Convention 1958.
Article V (2)(b) New York Convention.
Article 3.1 Scottish Short form Arbitration Rules 2012.
Article 3.2 Scottish Short form Arbitration Rules 2012.
Article 3.8 Scottish Short form Arbitration Rules 2012.
Article 5 UAA 2017.
Article 6 LCIA Rules.
Article 9B LCIA Rules.
Article 11 UNCITRAL Model Law.
Article 11(2) UNCITRAL Model Law.
Article 11(4) UNCITRAL Model Law.
Article 12 UNCITRAL model Law.
Article 13 ICC Rules.
Article 13(4) ICC Rules.
Article 13(5) & (6) ICC Rules.
Article 14 ICC Rules.
Article 18 Arbitration Act 1996.
Article 21(6) Arbitration foundation of South Africa.
Article 32(2) UNCITRAL Model Law.
Article 34(2)(b)(ii) UNCITRAL Model Law.
Article 35 UNCITRAL model law.
Article 1450 Code of Civil Procedure of France.
Paragraph 4 of the Draft Ethics Guidelines for Trustworthy AI: Working
Document for Stakeholders’ Consultation.
Part I IBA Guidelines on Conflicts of Interest in International Arbitration.
Section 8 Arbitration (Scotland) Act 2010.
Section 18 Arbitration Act 1996.
Section 69 English Arbitration Act 1996.
76 ARBITRATION: THE INT’L J. OF ARB., MED. & DISPUTE MGMT

11.4 WEBSITES

E. A. Adriano, The Natural Person, Legal Entity or Juridical Person and Juridical
Personality (2015), https://elibrary.law.psu.edu/cgi/viewcontent.cgi?article=
1117&context=jlia (accessed 21 Mar. 2022).
N. A.-S. Ahmed Mohammad Al-Hawamdeh, The Effects of Arbitrator’s Lack of
Impartiality and Independence on the Arbitration Proceedings and the Task of Arbitrators
Under the UNCITRAL Model Law (2018), https://www.researchgate.net/publica
tion/327320717_The_Effects_of_Arbitrator’s_Lack_of_Impartiality_and_
Independence_on_the_Arbitration_Proceedings_and_the_Task_of_Arbitrators_
under_the_UNCITRAL_Model_Law (accessed 28 Mar. 2022).
R. Calo, Artificial Intelligence Policy: A Primer and Roadmap (2017), https://ssrn.
com/abstract=3015350 (accessed 14 Feb. 2022).
F. Daso, A Harvard Law School Legal Tech Start Uo Uses AIto Settle Arbitration
(n.d.), https://www.forbes.com/sites/frederickdaso/2020/02/04/arbilex-a-har
vard-lawschool-legal-tech-startup-uses-ai-to-settlearbitrations/#4fc587da52c5
(accessed 15 Mar. 2022).
J. Dastin, Amazon Scraps Secret AIRecruiting Tool that Showed Bias Against
Women (2018), https://www.reuters.com/article/us-amazon-com-jobs-automa
tion-insight/amazon-scraps-secret-ai-recruiting-tool-that-showed-bias-against-
women-idUSKCN1MK08 (accessed Mar. 2022).
C. Dolinar-Hikawa, Beyond the Pale: A Proposal to Promote Ethnic Diversity
Among International Arbitrators (2015), http://sidley.com/-/media/publications/
tv124article17.pdf (accessed 28 Feb. 2022).
H. Eidenmüller & F. Varesis, What Is an Arbitration? Artificial Intelligence and the
Vanishing Human Arbitrator (2020), https://ssrn.com/abstract=3629145 (accessed
14 Feb. 2022).
S. Habib, The Use of Artificial Intelligence in Arbitration in Africa – Inevitable or
Unachievable? (2020), https://www.ibanet.org/article/E62B06F6-7772-458A-
A6E7-1474DB7136B5 (accessed 23 Mar. 2022).
A. Joshi, Replacing Arbitrators With Artificial Intelligence in International Arbitration
(2020), https://www.thearbitrationworkshop.com/post/replacing-arbitrators-
with-artificial-intelligence-in-international-arbitration (accessed 22 Mar. 2022).
Limited, Ebram International Online Dispute Resolution, https://www.ebram.
org/services.html (accessed 14 Feb. 2022).
Manhoor Waqar, The Use of AI in International Arbitration (n.d.), https://
papers.ssrn.com/sol3/papers.cfm?abstractid=3931233 (accessed 23 Mar. 2022).
J. McCarthy, What Is Artificial Intelligence (2007), http://jmc.stanford.edu/
articles/whatisai/whatisai.pdf (accessed 14 Feb. 2022).
 A STEP TOO FAR? 77

A. Misshra, Artificial Intelligence(AI) and Its Effects on Arbitration (2020), https://

taxguru.in/corporate-law/artificial-intelligence-ai-effects-arbitration.html
(accessed 22 Mar. 2022).
N. Butcher, M. Wilson-Strydom & Baijnath, Artificial Intelligence Capacity in
Sub-Saharan Africa: Compendium Report (2021), https://africa.ai4d.ai/wp-content/
uploads/2021/03/AI4D-Report%E2%80%94AI-in-SSA.pdf (accessed 23 Mar.
2022).
P. C. Nappert, The March of the Robots (2015), http://globalarbitrationreview.
com/article/1080951/the-march-of-the-robots (accessed 22 Mar. 2022).
J. W. Nelson, Machine Arbitration and Machine Arbitrators (2016), http://www.
youngicca-blog.com/machinearbitration-and-machine-arbitrators/ (accessed 22
Mar. 2022).
Young-Yik Rhim & Kyung Bae Park, The Applicability of Artificial Intelligence
in International Law (2019), http://dx.doi.org/10.14330/jeail.2019.12.1.01
(accessed 14 Feb. 2022).
N. Y. Section, Arbitration and Mediation (2018), http://www.nysba.org/pubs
(accessed 24 Feb. 2022).
L. C. Services, Development of an Online Dispute Resolution and Deal Making
Platform by Non-governmental Organisation (2019), https://www.legco.gov.hk/yr18-
19/english/panels/ajls/papers/ajls20190325cb4-665-3-e.pdf (accessed 14 Feb.
2022).
S. C. Statistics (n.d.), http://sccinstitute.com/statistics (accessed 28 Feb.
2022).
O. Tang, Think Arbi: Will Artificial Intelligence Help or Harm Arbitration? (2021),
http://arbitrationblog.kluwerarbitration.com/2021/12/22/think-arbi-will-artifi
cial-intelligence-help-or-harm-arbitration/ (accessed 23 Mar. 2022).

11.5 REGULATIONS
Regulation (EC) No 593/2008 of the European Parliament and of the Council of
17 Jun. 2008 (‘Rome I’).
Regulation (EU) No 1215/2012 of the European Parliament and of the
Council of 12 Dec. 2012 on jurisdiction and the recognition and enforcement of
judgments in civil and commercial matters (recast).

11.6 REPORTS

N. Butcher, M. Wilson-Strydom & Baijnath, Artificial Intelligence Capacity in Sub-

Saharan Africa: Compendium Report (2021).