Artificial intelligence in arbitrations: opportunities and challenges
In recent years, new technologies has made significant contributions in the area of dispute resolution, both in proceedings in the ordinary courts and in alternative methods, such as arbitration. Common examples of these advances are the use of digital means to perform acts that used to be "analogical", such as filing pleadings, issuing decisions, sending summons, and holding hearings, all of which have brought greater economy and efficiency to arbitral and judicial proceedings.
Even more recently, the emergence of artificial intelligence and the resulting automation of tasks using algorithm-based neural networks and machine learning has the potential to transform the relationship between legal procedure and technology. Earlier digital technologies still played an auxiliary role, serving as instruments to be used by legal professionals; artificial intelligence opens the way to replacing typically human analytical attributes and to producing material with enormous disruptive potential for proceedings in general, whether judicial or arbitral.
The Brazilian courts have already seen some of the impacts of this trend. The enormous database generated by the widespread digitalization of legal proceedings and public access to those proceedings (as a rule) has made it possible for the courts to develop big data tools like Athos, used by the Superior Court of Justice (STJ) to conduct preliminary screening of cases for frequently-litigated issues that may fall under the "repetitive appeals" system of binding rulings1, and project Victor, under development by the Supreme Federal Court (STF), which aims to optimize analysis of appeal admissibility2. These resources are not limited to judicial activity: the legal profession now has available to it intelligent programs and software to research precedents, locate assets, assess risks, and even draft pleadings3.
Arbitration faces an initial difficulty in implementing some of AI systems because of the lack of publicly-available information on arbitral proceedings: confidentiality of proceedings and awards is still the rule, despite recent efforts to increase transparency. On the other hand, arbitration is by its nature flexible, adaptable and oriented to parties' needs and market forces, which makes arbitration fertile ground for the application and development of AI-power tools, as long as the consensual nature of arbitral proceedings is respected, along with due process and other procedural guarantees.
AI tools can be used in various ways at different phases and by various actors in arbitral proceedings. There is a vast - and still largely unexplored - range of possibilities.
In a highly informative overview of the subject4, Professor Carlos Alberto Carmona mentions AI tools that support risk assessments by the parties, to analyze the viability of their claims and the chances of success; tools that help parties in the selection of arbitrators by filtering information to find the profiles most suitable for the resolution of their dispute; tools to aid in complying with the selected arbitrator's duty of disclosure and conflict of interest checks; tools that assist in drafting pleadings, submissions and briefs and in managing documents (which can be of considerable importance to the arbitrators, given the significant volume of documents usually produced by the parties); and still other tools designed to support the production of expert reports by legal and other specialists.
Implementing AI tools in arbitration faces certain challenges, however, especially in connection with the confidentiality of arbitrators' and attorneys' work, which must be preserved when using artificial intelligence systems. Ensuring confidentiality within the architecture of big data systems and machine learning is a complex task. There is no doubt, however, that artificial intelligence is here to stay in arbitration and offers opportunities as well as challenges.
NOTES
1 For more information, see SANSEVERINO, Paulo de Tarso; MARCHIORI, Marcelo Ornellas. O Projeto Athos de inteligência artificial e o impacto na formação dos precedentes qualificados no Superior Tribunal de Justiça. In: O Direito Civil na era da inteligência artificial. Rodrigo da Guia Silva e Gustavo Tepedino (coord.). 1 ed. São Paulo: Thomson Reuters Brasil, 2020.
2<https://portal.stf.jus.br/noticias/verNoticiaDetalhe.asp?idConteudo=471331&ori=1>. Acesso em 11.9.2023.
3 CABRAL, Antonio do Passo. Processo e tecnologia: novas tendências. In: Revista do Ministério Público do Estado do Rio de Janeiro nº 85, jul.set. 2022.
4 CARMONA, Carlos Alberto; VIEIRA, Vitor Silveira. Inteligência artificial e processo arbitral. In.: DIREITO, MERCADO JURÍDICO E SOCIEDADE: Estudos em comemoração aos três anos do grupo de jovens advogados Leading Young Lawyers. Disponível em <https://edisciplinas.usp.br/pluginfile.php/5829779/mod_resource/content/1/Intelige%CC%82ncia%20artificial%20e%20processo%20arbitral.pdf>. Acesso em 11.9.2023.
