Pooja
Damodaran*
“Our intelligence is what makes us
human, and AI is an extension of that quality.” – Yann Le
International arbitration has developed leaps and bounds
the past decade. International arbitration is borderless, flexible, efficient
and serves transnational justice. In the era of globalisation and technology
take over, there has been much speculation about not just process of
arbitration being technologically advanced but if the arbitrators can also be
replaced by robots. Artificial Intelligence (AI) is making its presence felt in
international arbitration but will the AI future be a realistic solution
to disputes or will there be no significant impact?
The future beholds
the answer.
In international arbitration, AI has been used for wide range of tasks including appointment of
arbitrators, legal research,
drafting, reviewing written
submissions, translating of documents, case management and document
organization, cost estimation, hearing arrangements, drafting procedural
orders, etc.
There are several studies that support computer programs
are better than humans in predicting the outcome of legal decision-making. For instance,
Jurimetria, the legal prediction technology tool belonging to Wolters Kluwer
provided analysis of millions of court decisions, information and the success rate of appeal lodged with the
Spanish courts. Siarelis robot is another example in this front, the Colombian
government used Siarelis in Superintendencia de Sociedades de Colombia to resolve
corporate disputes.
The settlement of disputes
by robots is undoubtedly causing great concerns. The main concern being ars boni et aequi. The
requirement imposed under arbitration laws give wide discretion to the parties to appoint arbitrators of
their choice. Most arbitration laws
require the arbitrator to be a natural person and have a full legal capacity.
Certain jurisdictions grant power to appoint robot arbitrators within the scope
of their definition of arbitrators. However,
certain jurisdictions such as England and France explicitly require arbitrators
to be natural persons.
Besides the debate if
it is permissible to appoint robotic arbitrators, there is also general
concern if the awards
generated by the robots can be equated
to man-made awards.
Robots generate awards based on rule-based algorithm
and machine-learning algorithms. The result generated
by the robot based on these algorithm are standard. The major drawback
in robot generated awards are the requirement of reasons. Article 31(2) of the
UNCITRAL Model Law for instance requires arbitrators to state reasons upon
which the award is based.
Where there is a requirement under the arbitration law
and/or the applicable arbitration rules to give reasons, the question
arises as to whether a failure to give reasons
for all or part of the decision
constitutes a valid ground for seeking to set aside the award.
Hence, the courts
generally set a low
standard for requirement to give reasons, however it is essential that the
arbitrators explain succinctly why they have reached their decision.
The reasons by algorithm are still not widely accepted
in international arbitration practice. There can be no definitive way to
understand how the algorithm reached its decision. These algorithm are human designed and thus, rudimentary reasoning of robotic
awards though accurate
is still not accepted as justice served.
Many tasks performed by the arbitrators such as legal ethics, evaluating issues and drafting awards
involve “people skills”, human interaction and understanding. There is no guarantee that a robot completely
resonates with the submissions of the counsels.
It can be concluded that idea of robot arbitrator is a big leap forward in terms of
technological advancement. There are some downsides to this AI tool. Robot generated awards are not legitimate
because justice is not a simple algorithm; it is human virtue. Justice needs
to be put to practice and needs a complex analysis
of situations and circumstances considering relevant facts, applicable law and render rationale solutions.
Even though robot justice is criticised for being
fallible and palatable, AI and international arbitration as whole does not
fail.
* LLM Candidate in International Commercial Arbitration Law (ICAL) in Stockholm
University. Email: : pooja.damodaren@gmail.com
[1] Yann
Lecun Professor, New York University.
[2] Robert H. Smit, The Future of Science and Technology in International
Arbitration: The Next Thirty Years, in The Evolution and Future of
International Arbitration p.365–78 (Wolters Kluwer 2016); Francisco Uríbarri
Soares, New Technologies and Arbitration, VII(1) Indian J. Arb. L. 84 (2018);
Gauthier Vannieuwenhuyse, Arbitration and New Technologies: Mutual Benefits, 35
J. Int’l Arb. 119–29 (2018); Sophie Nappert, The Challenge of Artificial
Intelligence in Arbitral Decision-Making, Practical Law UK Articles (4 Oct.
2018); Emma Martin, The Use of Technology in International Arbitration, in 40
Under 40 International Arbitration 337–48 (Carlos Gonzalez-Bueno ed., Wolters
Kluwer 2018).
[3] Roger Guimerà & Marta Sales-Pardo, Justice Blocks and
Predictability of U.S. Supreme Court Votes, (2011).
[4] “Supersociedades,
el primer juzgado de América Latina en contar con un robot asistente”,
Superintendencia de Sociedades” (Supercompanies, the first Latin American Court
to have a robot assistant, https://www.supersociedades.gov.co/Noticias/Paginas/2018/Supersociedades-el-primer-juzgado-de- Am%C3%A9rica-Latina-en-contar-con-un-robot-asistente.aspx
(Last visited: 12/03/2020).
[5] Digests
I.I.I., Nam, ut eleganter Celsus definit, ius est ars boni et aequi: law is the
art of the good and the equitable.
[6] Section 7 of the Swedish Arbitration Act: Any person who possesses
full legal capacity in regard to his actions and his property may act as an
arbitrator; Article 11(1) of the UNCITRAL Model Law provides: ‘No person shall
be precluded by reason of his nationality from acting as an arbitrator, unless
otherwise agreed by the parties’; The New York Convention does not impose any
requirement for arbitrators to be natural persons, and that Contracting States
that introduce such a requirement might arguably violate their obligations
under Article II of the New York Convention to recognise agreements to
arbitrate (Gary B. Born, International Commercial Arbitration, Vol. I, p. 293
(2nd ed., 2014).
[7] Section 26(1) of the English Arbitration Act, 1996; Art. 1450 of the
French Civil Code.
[8] Navigation Sonamar Inc. v. Algoma Steamships Limited (1994) XIX YCA
256; NJA 2009 p. 128 (Soyak II).
[9] The Nicomaquian Ethics, Aristoteles.
1 Comentarios
Great way of thinking and good luck for your future !
ResponderBorrar