Let us, for a while, think about a world where a computer, not a human, decides the outcome of a high-value dispute. While this might seem like something out of a science fiction movie, it’s rapidly turning into a reality in the field of international arbitration. Artificial Intelligence (AI) is changing the way arbitration is done, bringing efficiency, pace, and accuracy —but not without raising some critical questions. In this blog, let us explore the interesting interplay of AI in international arbitration by referring to real-life cases and establishing its benefits and limits.
The Rise of AI in Arbitration
Though the entry of AI into the legal field is not new, its application in arbitration has grown considerably in recent years. AI tools are now being used to automate administrative tasks, analyse huge amounts of data, and even predict arbitration outcomes. These advancements are reshaping the landscape of international arbitration.
“AI has the potential to take over repetitive tasks, allowing lawyers to focus on strategy and advocacy,” says Michael McIlwrath, Global Chief Litigation Counsel at GE Oil & Gas.
Here are a few benefits that can be seen as an outcome of AI usage in arbitration:
- Automation for Better Efficience
No other tool comes close to what AI can do when it comes to streamlining administrative tasks in arbitration. From managing documents to scheduling hearings, AI tools can handle huge amounts of data with accuracy and speed.
Think of AI as a tireless librarian, who diligently organizes a vast array of books in a library with the precision of a Swiss watchmaker.
Case Study: The International Chamber of Commerce (ICC) has utilized AI in over 500 cases under its Expedited Procedure Provisions since 2017. AI-driven document review and e-discovery have significantly reduced the time and cost associated with these processes, thereby enabling faster resolutions.
Document management systems and e-discovery software are among the AI tools that can quickly review and categorize a large number of documents, saving time and reducing costs. That way, all involved parties can save significant expenses in arbitration.
- Predictive Analytics for Better Decision-Making
The ability of AI to analyze large amounts of data can help us to identify trends and patterns, which were difficult, if not impossible, earlier. This is where predictive analytics comes to assist. Predictive analytics is a tool that can indicate possible outcomes from analyzing past data, thus offering a more informed decision-making capability. The arbitrating parties can not only understand the strengths and weaknesses of their positions but also undertake more informed negotiations.
Case Study: The case, Pyrrho Investments Ltd. v. MWB Property Ltd marked the first time when predictive coding was allowed for e-discovery in the UK. This considerably reduced the time required for document review and lowered labour costs. This even made the process more affordable for the involved parties.
- Improved Accuracy
AI can help in minimizing human errors in document review and data analysis. This ensures that arbitration decisions are based on accurate and comprehensive information.
Case Studies: The US has employed AI-powered tools such as COMPAS (Correctional Offender Management Profiling for Alternative Solutions) for data collection, risk assessment, and decision support.
The Supreme Court of India’s SUPACE (Supreme Court Portal for Assistance in Courts Efficiency) system is a prime example. SUPACE, an AI-driven system, aids judges by summarizing legal documents and providing insights. Despite initial concerns about data privacy and security, the system has proven to enhance the accuracy and efficiency of judicial processes.
Prior to this, the Apex Court had launched an indigenously engineered neural translation tool, SUVAS (Supreme Court Vidhik Anuvaad Software), a machine-assisted translation tool trained by AI to quickly and accurately translate judicial orders and rulings from English to local languages.
Smart Solutions Meet AI Advantage
In addition to the above benefits, AI has been put to use in various aspects of arbitration.
- Natural Language Processing (NLP) in AI systems is being used to interpret and understand legal language, thereby helping to identify prior case laws and draft legal documents.
- AI can be used by parties to estimate and plan a budget by analyzing data from previous cases.
- AI can be used to optimise case strategies for an ongoing case. By analyzing the strength and weaknesses as identified in the evidence and arguments, AI can suggest the optimal strategy in the way forward.
AI can review and scan through arbitrators’ professional connections and identify any conflict of interest. This can ensure impartial proceedings during arbitration.
Now, let’s move from the benefits of AI to talk about some of its drawbacks. AI is much like that star student in our law class who always has the right answers and never misses a lecture. But then, sometimes gets overly confident and starts making silly mistakes, like citing a tort case in a contract law exam. Similarly, AI can interpret large data, draft documents faster than we can say “res ipsa loquitur,” and even make predictions like a pro. But, just like how you’d never let that eager law student argue your real case without supervision, you shouldn’t let AI operate without human oversight in arbitration.
Defining the Limits of AI in Arbitration
Though AI offers many benefits, it is important to keep human oversight in arbitration. Relying too much on AI could weaken the human judgment required in legal decision-making. Judges and arbitrators must make sure that AI remains a tool to aid their work rather than replace their critical thinking, experience, and expertise.
“While AI can assist in minimizing errors in numerical calculations, it cannot replace the human capacity for compassion that is crucial in judicial reasoning,” remarks former Justice DY Chandrachud of the Supreme Court of India.
Here is a look at few limitations of using AI in international arbitration:
- Maintaining Human Oversight
Although AI offers several advantages, it often struggles to handle the detailed and complex nature of legal reasoning. It is important that Judges and arbitrators retain the authority to take the final decision, thus ensuring fairness and accountability. The Supreme Court of India’s SUPACE system, designed to assist judges with AI-driven insights, emphasizes the need for judges to critically evaluate AI outputs rather than depend on them entirely.
Case Study: In the arbitration case of Occidental Petroleum v Ecuador, inaccurate translations arising from AI software resulted in loosing of a claim for one of the parties. The case highlights that verifying AI output is critical to avoid misinterpretation.
- Ensuring Fairness and Impartiality
AI algorithms may reflect the biases in their training data, which can lead to unfair outcomes. In a high-profile arbitration involving a multinational corporation, AI’s recommendations were seen to favour the larger entity due to biased data inputs. This case emphasized the need for creating frameworks that can detect and redress biases in AI systems.
Book Reference: In the book “Weapons of Math Destruction,” the author Cathy O’Neil, the author explores how biased algorithms can continue and even worsen existing social inequalities. In his book, “Human Compatible,” Stuart J. Russell argues for the development of AI systems that align with human values and the need for ethical guidelines to govern AI use.
It is important to establish clear guidelines that seem necessary to maintain transparency in AI methodologies that are crucial to ensure accountability.
- Protecting Confidentiality
Data privacy is another critical concern. AI tools must adhere to stringent data protection measures to safeguard sensitive information. The Supreme Court of India faced significant concerns about data privacy and security while deploying its SUPACE system. Just as a fort requires strong gates and vigilant guards, there needs to be robust security protocols to prevent unauthorized access and breaches.
In recent arbitrations managed by the Silicon Valley Arbitration and Mediation Center (SVAMC), implementing AI is guided by established guidelines issued in April 2024. The guidelines focus on preventing unauthorized access and data breaches in the use of AI tools. Without stringent protection of confidential data, there could be a compromise on the fairness of the arbitration process.
Future of AI in Arbitration
With the advancement of AI, the future of international arbitration will likely involve more often to ensure an increase in efficiency and an enhanced decision-making capability.
However, it is crucial to set up clear limits so that AI supports the process of arbitration and not weaken it. If we address ethical concerns and keep human oversight, we can make the most of AI’s potential while maintaining the integrity and fairness of international arbitration.
Unquestionably, tomorrow’s arbitrations will have more room for AIs. As we embrace this technological advancement, we must ensure that AI serves as a tool to enhance human capabilities, not substitute them completely.
Conclusion: The Road Ahead
As we step into this new era, the question remains: How will we balance the power of AI with the wisdom of human judgment? The future of international arbitration depends on finding the right equilibrium between these two forces.
By Bhaskar Dutta, 2nd Year LL.B (General), Faculty of Law, The Maharaja Sayajirao University of Baroda, Vadodara
